Tumour plasticity and tumour microenvironment interactions as potential immunologic targets for pancreatic cancer treatment

Introduction

Pancreatic ductal adenocarcinoma (PDAC) is an aggressive malignancy with the worst prognosis among all the other major cancer types, having an overall five year survival rate for all stages of only 13% (1). Despite improvements in systemic therapies, it has become the fourth leading causes of cancer death in the United States (1-3). The development of adjuvant chemotherapy for resectable PDAC and neoadjuvant therapy for borderline resectable cancer has significantly improved the outcomes of patients with early stage PDAC (2,4,5). However, most patients with more locally advanced and or metastatic disease PDAC have a median survival of barely 8–12 months even with treatment and no long-term survivors, whilst targeted therapies have so far been disappointing (3,6). Immunotherapy uses the power of the body’s own immune system to suppress and eliminate tumour cells, which has been shown to be a promising approach in several solid malignancies such as melanoma and breast cancer has so far shown only a poor response to immune therapy (7-9).

PDAC tumour cells are well-embedded in a tumour-suppressive ecosystem, surrounded by an intense desmoplasia in a complex tumour microenvironment (TME) which can be up to 90% of the tumour in some cases (10). The desmoplastic stroma is composed of extracellular matrix (ECM), cancer-associated fibroblasts (CAFs), endothelial cells, blood and lymphatic vessels and nerves, and most importantly a variety of largely dysfunctional immune cells (11,12). The tumour immune microenvironment (TIME) includes tumour infiltrating lymphocytes (TILs) comprising of minimal to moderate numbers of CD4⁺ and CD8⁺ T cells, more peripheral tertiary lymphoid structures (TLSs) with T and B lymphocytes, and dendritic cells (DCs) which are compartmentalised for the generation of local antitumour immunity (13-16). Chronically activated, tumour-reactive T cells exposed to fibroblast-derived TGF-β act as TLS organisers by producing the B cell chemoattractant CXCL13, resulting in B cells proliferation and differentiation into plasma cells (14). Galon and Bruni have identified tumour infiltration by CD8⁺ and T-helper1 cells and TLS along to be prognostically favourable whilst infiltration by M2 macrophages is prognostically poor (17). Despite the immune-cold immunosuppressive TME in PDAC, experiments in murine models and early-stage trials have indicated the possibility of switching a cold state into a hot immune one, improving the possibility of clinically meaningful immunotherapy. With whole tumour and single-cell sequencing in both PDAC tumour cells and its surrounding TME, we can sketch an all-in-view profile of tumour/TME/CAF subtypes and found to transition with tumour progression over time and in response to chemotherapy and immunotherapy (18-21). The trajectory of the tumour/TME/CAF subtype interaction and transition may provide new perspectives in immunotherapy (Figure 1).

Figure 1 TME/CAF/tumour subtypes and their interactions through an immune-cold state (deserted TME/basal subtype) to an immune-hot state (reactive TME/classical subtype) trajectory. TME, tumour microenvironment; CAF, cancer associated fibroblast; NK, natural killer; DC, dendritic cell; Th1/2, T helper cell 1/2; Treg, regulatory T cell; PSC, pluripotent stem cell; MDSC, myeloid-derived suppressor cell; TAM-M1/M2, tumour-associated macrophage type 1/2; iCAF, inflammatory cancer-associated fibroblast phenotype; myCAF, myofibroblast cancer-associated fibroblast phenotype; apCAF, antigen-presenting cancer-associated fibroblast phenotype; IL, interleukin; αSMA, α-smooth muscle actin; MHC-I/II, major histocompatibility complex I/II; TGF-β, tumour growth factor-β; PDFGF, platelet-derived growth factor; CTGF, connective tissue growth factor; CD, cluster of differentiation; CTLA-4, cytotoxic T-lymphocyte associated protein 4; FOXP3, forkhead box protein 3; TCR, T cell receptor; CXCR3, C-X-C motif chemokine receptor 3; CCR5, C-C motif chemokine receptor 5; IFNG, interferon-γ; GATA3, GATA binding protein 3; STAT3/6, signal transducer and activator of transcription 3/6; HIFα, hypoxia inducible factor-α; IDO, indoleamine; NKG2D, natural killer cell granule protein 2D; TIGIT, T cell immunoreceptor with Ig and ITIM domains; FIZZ1, “found in inflammatory zone 1” protein; TLR, toll-like receptor; iNOS, inducible nitric oxide synthase; Fas/FasL, fas cell surface death receptor/ligand; PD-1/PD-L1, programmed cell death protein 1/ligand 1; ARG1, arginase 1; ROS, proto-oncogene 1; CSF, colony-stimulating factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; ISG15, interferon-stimulated gene.

Current immunotherapy strategies

The general strategies for PDAC immunotherapy to turn the cold state into an immune-hot one include the following (3):

• To enhance neoantigen presentation and activation antigen-presenting cells [APCs; mRNA vaccines, peptide vaccines, granulocyte-macrophage colony stimulating factor (GM-CSF) vaccines];
• To increase the number of effector T cell population in the tumour by stimulating recruitment, activating T cells by utilizing immune checkpoint inhibitors(ICIs) or costimulatory factors such as CD137 [tumour necrosis factor receptor superfamily member 9 (TNFRSF9), 4-1BB] or CD40 agonists or by direct administration of chimeric antigen receptor (CAR) or T cell receptor (TCR) T cells;
• And to modulate the immunosuppressive elements in TME such as the M2 tumour-associated macrophages (TAMs; CSF1R inhibitor, and FAK inhibitor), myeloid-derived suppressor cells (MDSCs; CCR2 inhibitor) and CAFs (vitamin D receptor agonist). A selection of completed clinical trials on PDAC immunotherapy were listed in Table 1.

ICIs

A number of immune checkpoint regulators are involved in down-regulating immune activity in pancreatic cancer.

Programmed cell death protein 1 (PD-1) is cell surface receptor inducibly expressed on T cells, B cells and myeloid cells. It is activated by its ligand (PD-L1) which down-regulates immune activity and promoting self-tolerance but also prevents cancer cell killing when expressed by the tumour cells (33).

Cell surface receptor cytotoxic T-lymphocyte associated protein 4 (CTLA-4) is constitutively expressed on regulatory T cells (Tregs) and upregulated in conventional T cells in tumours. The CD28/CTLA-4 co-stimulatory pathway has a crucial role in regulating T-cell activation and tolerance. Downstream signalling is activated when bound to costimulatory molecules CD80 or CD86 expressed on the cell membrane of APCs. Whereas binding with CTLA-4 inhibits T-cell effector function, interaction of the ligands CD80/CD86 with CD28 on T-cells results in enhanced and sustained T-cell activation (34).

T cell immunoreceptor with immunoglobulin and ITIM domain (TIGIT) is upregulated on activated T cells, natural killer cells, and regulatory T cells. TIGIT binds to two ligands, CD155 and CD112 (nectin-2), that are expressed by tumour cells and APCs (35,36). Clinically advanced PDAC tumours have high levels of T cell exhaustion and correlate with TIGIT expression.

V-domain immunoglobulin suppressor of T cell activation (VISTA) in PDAC is predominantly found on CD68⁺ macrophages, activation of which decreases CD8⁺ T cell responses (37).

CD39 and CD73 immune checkpoint mediators

Adenosine triphosphate (ATP) can be released at high levels from malignant cells, provoking inflammation by purinergic signalling promoting anti-tumour responses. Hydrolysis of extracellular ATP by the membrane-bound ectonucleotidases generates CD39 and CD73 immunosuppressive adenosine (38).

T cell immunoglobulin and mucin domain-containing protein 3 (TIM3)

This is s a transmembrane protein found on multiple immune cells including macrophages, as well as NK, dendritic, myeloid-lineage, and T cells, with both inhibitory and co-stimulatory functions, depending on the cell type and molecular interactions. TIM3 is expressed on exhausted T cells functioning as a key checkpoint inhibitor. Blocking both TIM3 and PD-1 results in improved T cell function with an anti-tumour response in patients with advanced PDAC (39).

Lymphocyte-activation gene 3 (LAG-3), is expressed on activated T cells, natural killer cells, B cells and plasmacytoid DCs. Its main ligand is major histocompatibility complex (MHC) class II, to which it binds with higher affinity than CD4, down-regulating immune activity (40).

Dysfunctional and exhausted CD8⁺ T cell will express PD-1^Hi, CTLA4⁺, TIM3⁺, LAG3⁺ and TIGIT⁺ (41). The exhausted TIME may also be characterised by higher expression of EOMES (eomesodermin) and GZMK (Granzyme K): EOMES is a lineage-defining transcription factor in IFN-γ/IL-10 coproducing Treg1-like cells and they also express GZMK, but not CD40L and IL-7R in contradistinction to all other CD4⁺ T-cell subsets, including conventional cytotoxic CD4⁺ T cells (42). In general ICIs to PD-1, PD-L1 and CTLA-4 have failed to provide a clinically significant benefit in PDAC, which is ascribed to dysfunctional T cells, paucity of neoepitopes, and the activation of alternative immunosuppressive immune checkpoint, such as TIGIT and VISTA. In accordance with other tumour types however, PD-1 inhibitors are effective in pancreatic cancers with high burden of microsatellite instability (MSI-high) but is present in less than 1% of cases. In a retrospective study of nine patients (8 had single agent pembrolizumab, and one received ipilimumab plus nivolumab) seven (77%) had a RECIST response (43). Pathology-assisted deconvolution of spatial transcriptomic data in PDAC tissues revealed coordinated expression of TIGIT in exhausted and regulatory T cells and Nectin in tumour cells whilst chemo-resistant samples showed enrichment of inflammatory CAFs upregulate metallothioneins (44). In experimental PDAC models the CD155/TIGIT ligand-immune receptor axis mediates immune evasion, and when targeted in combination with PD-1 blockade and CD40 agonism induced anti-tumour immunity (35). Thus, blocking more than one immune checkpoint pathway is a possible strategy but poses a number of hurdles including the risk of cumulative toxicity.

Another important escape mechanism is copy-number alterations (CNAs), which alter the dosage of multiple linked genes, and show recurrent patterns associated with clinical outcomes. Loss of chromosome 9p21.3 is the most strongly linked to poor prognosis and is found in around 60% of PDAC tumours. Deletion of 9p21.3 disables both cell-intrinsic and cell-extrinsic tumour suppression and is associated with increased resistance to ICIs. This is s a transmembrane protein found on multiple immune cells including macrophages, as well as NK, dendritic, myeloid-lineage, and T cells, with both inhibitory and co-stimulatory functions, depending on the cell type and molecular interactions. TIM3 is expressed on exhausted T cells functioning as a key checkpoint inhibitor. Blocking both TIM3 and PD-1 results in improved T cell function with an anti-tumour response in patients with advanced PDAC. The type-I interferon (IFN) genes and the CDKN2A/B tumour suppressor genes are co-located on the 9p21.3 locus. The type-I IFN-I genes consist of IFNα, IFNβ, IFNδ, IFNϵ, IFNκ, IFNω, and IFNτ, and the type II IFN comprises the IFNγ gene. Deletion of the 9p21.3 locus can result in either loss of CDKN2A alone or be co-deleted with the IFN gene cluster. In the TME, type-I IFN’s are required for robust immune surveillance and adaptive immune responses either directly or by stimulating immune effector T-cells, NK cells and macrophages. Loss of the 9p21.3 CDKN2A/B and IFN gene cluster contributes to resistance of immune checkpoint blockade therapy (45) (19a). Loss of chromosome 9p21.3 therefore has potentially important decision making for patient immune specific therapy selection.

Cancer vaccines

Cancer vaccine immunotherapy aims to prime new or expand the existing TCR repertoire of anti-tumour T cell responses (46). Various vaccine trials have shown induction of induction of T cell immunity GM-CSF-secreting allogeneic pancreatic tumour cells (GVAX) but have previously failed to demonstrate a clinical benefit in patients with advanced or PDAC early PDAC as adjuvant therapy (Table 1). GVAX is currently under investigation in combination with urelumab (an anti-CD137 agonist expressed on T-cells) and nivolumab (an ICI) in patients with resectable PDAC (NCT02451982).

Peptide-based vaccine trials targeting mtKRAS include the AMPLIFY-7P trial, investigating a lipophilic modified RAS peptide vaccine called ELI-002, which contains seven lipid-conjugated peptide-based KRAS antigens (G12D/R/V/A/C/S/D) (Amph-Peptides) plus a lipid- conjugated immune-stimulatory oligonucleotide (Amph- CpG-7909) (NCT05726864) (47). Alternative approaches involve mRNA technology for generating vaccines to mtKRAS including mRNA-5671, which encodes G12D, G12V, G13D, and G12C-specific peptides and are presently under investigation in combination with pembrolizumab (NCT03948763) (47).

A novel approach is to use an individualized neoantigen vaccine based on uridine mRNA-lipoplex nanoparticles (cevumeran). In a phase I trial of adjuvant autogene therapy with cevumeran (a maximum of 20 neoantigens per patient), atezolizumab (anti-PD-L1), and mFOLFIRINOX, from 16 patients treated, high-level neoantigen-specific T cells were induced in eight patients, and in four of these the responses targeted more than one vaccine i.e., they were polytopic (48). The enlarged vaccine-induced T cell population consisted of 10% of all the circulating T cells, which was further expanded with repeat vaccination and included neoantigen-specific CD8⁺ T cells with multiple effector functions. Responder patients with vaccine-induced T cells had a greatly increased median recurrence-free survival rate compared to non-responder patients (48).

Adoptive T cell transfer (ACT)

T cells have receptors comprised of heterodimers of either α and β chains, or γ and δ γδT cells normally do not express the co-receptors CD4 and CD8, and account for, on average, 4% of human peripheral blood T cells. ACT requires extraction functional αβT cells from the patient’s tumour tissue or blood, engineered to better recognise tumour cells, amplificated in vitro and infusion back into the patient.

Tumour-infiltrating lymphocyte (TIL) therapy

TILs are already primed against the cancer cells but require both expansion (with IL-2) and exclusion of γδT cells and Tregs. Just prior to receiving the TIL infusion patients undergo several rounds of high-dose non-myeloablative lymphodepleting chemotherapy eliminating immunosuppressive cells, including Tregs cells and MDSCs (such as cyclophosphamide followed by fludarabine) and subsequent to TIL infusion the patients are also given an abbreviated high-dose course of IL-2 (49). Lifileucel (Amtagvi) is the first cancer treatment using TILs approved by the Food and Drug Administration (FDA) in February 2024 against melanoma.

In an ongoing MD Anderson trial (NCT03610490), CD8⁺ predominant ex vivo TIL production with IL-2 and agonistic stimulation of CD3 and 4-1BB antibodies used in early TIL culture observed a best response of prolonged stable disease in one of five patients with PDAC lasting 17 months (50). Infusion product phenotypes were generally enriched for CD8⁺ rather than CD4⁺ TILs, which was expected because of the use of the 4-1BB agonistic antibody used in manufacturing process. Four effector memory subsets were found, mostly CD27⁻CD28⁻, which was associated with a differentiated cytotoxic phenotype. There was low expression of checkpoint markers including PD-1, CTLA-4, LAG3, TIM3 and TIGIT across the bulk CD8⁺ and CD4⁺ populations indicating strongly activated CD8⁺ TILs without exhaustion attributes showing CD69 expression and high expression of CD39 (50).

T cell receptor T cell (TCR-T) therapy

TCR-T consists of genetically engineered T cells, modified to express a receptor directed against one or more tumour antigens. The TRC complex comprises of a TCR constructed of an α and β chain heterodimer linked to CD3 subunits in the external cell membrane and closely associated with a CD4 or CD8 co-receptor plus lymphocyte-specific protein tyrosine kinase (Lck; a non-receptor Src family kinase). There are four surface membrane chains CD3γ, CD3δ, and two CD3ε chains and two internal membrane CD3ζ chains. The CD3 co-receptor is involved in activating both the cytotoxic T cell (CD8⁺ naive T cells) and T helper cells (CD4⁺ naive T cells). TCR-T has potentially wider selection of tumour targets, as it recognises intracellular proteins presented as antigens by MHC-I. On-target off-tumour effects can be fatal if the target is co-expressed in non-tumour tissue. MHC class I (HLA-A, HLA-B, and HLA-C) present antigens from tumour cells to CD8⁺ co-expressing cytotoxic T cell TCRs triggering programmed cell death of the tumour cell by apoptosis. Each HLA can present a different set of antigens and different TCR-T products can target the same antigen.

Antigen selection is the key to successful TCR-T cell therapies: the ideal antigen would be expressed selectively and homogeneously in tumour cells and generate epitopes presented on MHC class I molecules on the TCR-T cell surface (51,52). The two main classes are (I) tumour-associated antigens (TAAs) specifically tissue differentiation antigens and cancer germline antigens or (II) tumour-specific antigens (TSAs), specifically TSAs and mutation-associated neoantigens (Figure 2).

Figure 2 Mechanisms and differences between TCR-T and CAR-T therapies. TCR-T targets only on MHC-presented antigens, which broadens the potential targets from surface to intracellular proteins. CAR-T works in an MHC-independent way by directly binding to tumour cell surface antigens. The efficacy of TCR-T is directly influenced by MHC expression in tumour cells. Both treatments have on-target off-tumour risks. TCR-T, T cell receptor T cell; CAR-T, chimeric antigen receptor T cell; TCR, T cell receptor; MHC, major histocompatibility complex; IL, interleukin; STAT 3/5, signal transducer and activator of transcription 3/5; JAK, Janus kinase; VL, variable light; VH, variable heavy.

Leidner et al. treated a patient with progressive metastatic PDAC with a single infusion of 16.2×10⁹ autologous T cells, genetically engineered to clonally express two allogeneic HLA-C*08:02-restricted TCRs targeting mtKRAS G12D expressing tumours (52). The patient had overall partial response of 72% with regression of visceral metastases which was ongoing at 6 months. The engineered T cells constituted more than 2% of all the circulating peripheral blood T cells 6 months after the cell transfer. There are now several ongoing phase I/II clinical trials evaluating TCR-T directed against mtKRAS in PDAC including G12V (NCT04146298, NCT03190941), KRAS G12D (NCT03745326), TP53 (NCT05877599), and mesothelin (NCT04809766) (47).

Chimeric antigen receptor T cell (CAR-T) therapy

CARs are genetically engineered T cells that contain a surface membrane single-chain variable fragment (scFv) recognition domain derived from an antibody comprising a variable light (VL) and variable heavy (VH) chain fused to a transmembrane domain. To date, five generations of CARs have been developed (Figure 2) (53).

• The first-generation CAR consists of scFv fragment connected to a CD3ζ chain, acting as the transmembrane signalling domain to mediate antigen-dependent activation. Signalling is initiated by Lck-mediated phosphorylation of immuno-tyrosine activation motifs (ITAMs) within the cytoplasmic domains of CD3ζ.
• The second-generation CAR adds a single co-stimulatory molecule (CD28, 4-1BB, OX40, CD27, or ICOS) to enhance T cell responses.
• The third generation CARs now include two co-stimulatory domains linked to the to the CD3ζ chain improving anti-tumour efficacy.
• The fourth generation CARs were further improved to have an interleukin inducer, regulated by Ca²⁺ dependent NFAT (nuclear factor of activated T-cells) dephosphorylation which leads to cytokine release from the CAR-T cells. The fourth generation CAR-Ts are also called TRUCKs referring to T cells redirected for antigen-unrestricted cytokine-initiated killing. The term armoured CAR-T describes the genetic engineering strategy to encode for secretion of cytokines, modulation of cytokine function, or secretion of antibody-like proteins.
• The fifth generation is another armoured version CAR-T developed to enhance proliferation and anti-tumour activity by insertion of IL-2Rb, inducing antigen dependent activation of the JAK-STAT pathway.

There are now six different FDA-approved commercial CAR-T cell therapies but none so far for solid malignancies. Although phase I/II studies are now ongoing in solid malignancies and PDAC, the challenges are considerable (Table 1). Circulating CAR-T cell levels in patients with solid tumours tend to be rather low and data are limited as to whether they can be increased using lymphodepletion. Steven Albelda recently concluded that CAR-T cell therapy success for solid tumour may not be optimal if modelled on approaches that were successful in patients with haematological cancer which depend on engineered T cells with the least effector-like and most memory-like or stem cell-like phenotypes, that initially traffic to bone marrow and lymph nodes and with long-term persistence (54). A better approach for solid tumours might involve administering repeat doses of highly active effector-like cells that can traffic more efficiently to tumours with repeat administration as they lose function. In order to deal with tumour plasticity, the conditions during ex vivo T cell expansion can be modified, use of a CD28 cytoplasmic domain may protect from TGFβ-induced immunosuppression, along with TME activation and multiplexing using mRNA. Approaches designed to generate CAR-T cells in situ could be especially attractive (54).

Inflammation and myeloid cell-based therapies

Chronic inflammation (pancreatitis) and early stage PDAC are both characterised by an acquisition and accumulation of KRAS mutations. Chronic inflammation provides a microenvironment that contains activated leukocytes including macrophages, DCs, neutrophils, mast cells and T cells which secrete a wide variety of cytokines and chemokines providing strong mitogenic signals to the epithelial cells and CAFs. Local release of enzymes will also lead to increased levels of reactive oxygen species, NFkB, and Cox-2 potentially linking inflammation with initiation to ADM and PanINs followed by progression. PDAC cells utilize inflammation to establish a tumour-permissive TME with an immunosuppressive propensity. Therapeutic targets for therapies, include the secreted IL-1β, IL-6, IL-8, CCL2, CCL5, and TNFα ligands as well as the infiltrating inflammatory cells. Therapies include targeting MDSCs and TAMs using for example blockade of CSF-1R, CXCR2 and CCR2 to reverse the TME immune-cold state (NCT02732938) (55-57). Preclinical studies showed that disrupting CCL2-CCR2 axis improved chemotherapy efficacy165 (58). A combination of CCR2 antagonist and FOLFIRINOX was utilized in a phase I trial and showed promising results with borderline resectable patients, while CCR2 plus gemcitabine with nab-paclitaxel caused high incidence of pulmonary toxicity (29,59). Another phase Ib trial of CCR2 and CCR5 dual antagonist in combination with chemotherapy is in process (NCT03184870).

Pancreatic cancer stem cell (PCSC)-based therapies

PCSCs are a major player in tumour plasticity, with the capacity for self-renewal, proliferation after drug treatment, and differentiation into divergent cell phenotypes (60). The programmable feature of cancer stem cells exposes them to potential therapeutic targeting. PCSCs comprise less than 1% of all PDAC cells, with overexpression of multiple stem cell and pluripotency markers including CD133, CD44, CD24, ESA and c-Myc, and activation of self-renewal pathways such as Sonic hedgehog (SHH), Wnt, NOTCH, and BMI-1 (61-63). PCSCs are chemoresistant and are negatively associated with patient survival (61,63,64). PCSCs are also involved in driving PDAC metastasis. CD133⁺CXCR4⁻ PCSCs may be classified as a stationary phenotype with the ability to promote the primary tumour, whilst the migratory CD133⁺CXCR4⁺ phenotype, may promote both tumour initiation and metastasis (65). Stem cell based anti-cancer therapy includes inhibiting specific PCSC signalling pathways such as Notch, Hedgehog, Wnt, Hippo, JKK/STAT, Nodal/Activin and KRAS (66). Novel approaches to targeting the pancreatic cancer are also based on naïve stem cells, anti-cancer drug loaded stem cells, genetically engineered stem cells and exosomal miRNA released by stem cells (67).

Stroma-based therapies

Focal adhesion kinase (FAK) activity in neoplastic PDAC cells is an important regulator of the fibrotic and immunosuppressive TME, and is upregulated in 80% of pancreatic cancer tissues, and is correlated with high levels of fibrosis and poor CD8(+) cytotoxic T cell infiltration (68). Preclinical studies have found that FAK inhibition led to decreased Treg cells, MDSCs and TAMs in the TME, with increased survival in mice model. Furthermore, FAK inhibitor plus gemcitabine-based chemotherapy also showed improved survival. FAK in combination with a PD-1 inhibitor scheme is ongoing (NCT04331041, NCT03727880, NCT02758587) (69,70).

Agonistic immunotherapies

One of the tumour evasion mechanisms is the deficiency of neoantigen presenting through MHC-1. Agonists of APCs including DCs, B cells, and macrophages can largely improve neoantigen recognition and activate downstream immune response. APCs become activated when CD40 binds with CD40L on CD4⁺ T-helper cells. CD40-agonistic antibodies targeting DCs targeted can shift TAMs from the M2 (tumour promoting) to the M1 (anti-tumour) subtype and active effector T cells. The NCT03214250 randomized phase II trial evaluated the efficacy of anti-PD-1 antibody nivolumab and/or the CD40 agonistic antibody sotigalimab with gemcitabine plus nab-paclitaxel in 105 patients with first-line metastatic PDAC (71). The results showed a higher 1-year overall survival for the nivolumab arm but not the sotigalimab arm. In the nivolumab arm was correlated with a less suppressive TME and higher numbers of activated (HLA-DR+) non-naive CD4⁺ and CD8⁺ antigen-experienced circulating T cells, whilst survival in the sotigalimab arm was correlated with greater intra-tumoral CD4⁺ T cell infiltration and circulating differentiated CD4⁺ T cells and APCs (72). A further clinical trial is testing the effect of CD40-agonistic antibody therapy with or without nab-paclitaxel and gemcitabine in the neoadjuvant setting (NCT02588443).

In a phase I/II study in metastatic PDAC using an oncolytic adenovirus with transgenes encoding TMZ-CD40L and 4-1BBL (LOAd703), which lyses cancer cells selectively, and combined with gemcitabine with nab-paclitaxel found that 8 (44%) of 18 patients had an objective response associated with increased CD8⁺ effector memory cells and adenovirus-specific T cells in 15 (94%) of 16 patients assayed (NCT02705196) (73).

PDAC tumour/TME/CAF heterogeneity and plasticity

Pancreatic cancer is characterized as a malignancy with high TME heterogeneity from transcriptional and morphological perspectives (10-12,18-21). In contrast, the immunosuppressive features in PDAC tumours seem to be a universal hallmark throughout tumour progression, hence the diverse immunotherapies have limited clinical efficacy (21). Personalized immune treatment according to PDAC tumour heterogeneity may be a possible strategy to consider.

Tumour initiation, progression and plasticity

Pancreatic cancer is initiated by the accumulation of somatic mutations evolving from early polyclonal and multifocal ductal pancreatic intraepithelial neoplasia (PanIN) lesions and intraductal papillary mucinous neoplasms (IPMNs) (10,74,75). An alternative pathway requires acinar cells undergoing acinar to ductal metaplasia which in the presence of a KRAS mutation in cooperation with the alarmin cytokine IL drives the metaplasia to PanINs by epigenetic mechanisms (75).

Cell plasticity refers to the ability to acquire new phenotype while tumour initiation and development, as a response of transcriptomic changes, drug treatment or as an adaption to surrounding environment (60,76). The acquisition of PDAC plasticity can happen in multiple stages during tumorigenesis. Once the tumours are generated, they show a high degree of heterogeneity and plasticity, which are considered important drivers of the TIMEs (10,19,46,74,77).

Intrinsic plasticity drivers

Mutational background

The most common single gene drivers involved in the malignant pathogenesis of PDAC are mutations and other gene alterations occur in the KRAS oncogenic driver in 90%, and the suppressor genesTP53 in 70%, CDKN2A in 60%, and in SMAD4 in 40% (10,74). Less common alterations in 5–10% of tumours are found in TGFBR2, the SWI/SNF subunit of ARID1A, the histone demethylase KDM6A, the histone H3K4 methyltransferase MLL3, the RNA-binding motif-10 regulating alternative splicing RBM10, the transcriptional corepressor BCORL1, and the roundabout guidance receptor 2 ROBO2, limiting stromal T-cell infiltration (10,74). In 80–100% of tumours key signalling and other canonical pathways affecting at least one gene in these pathways are KRAS signalling, DNA damage control, regulation of G1/S phase transition, TGFβ signalling, apoptosis, Hedgehog signalling, homophilic cell adhesion, integrin signalling, c-Jun N-terminal kinase signalling, regulation of invasion, small GTPase dependent signalling, and Wnt/Notch signalling. Less commonly affected in 5–25% of tumours are genes involved in histone modulation, SWI/SNF ATP-dependent chromatin remodelling complexes, RNA processing, and the ROBO/SLIT pathway (10,74).

Pancreatic acinar and epithelial cells with activated KRAS mutations can be triggered by inflammatory events, thereby gaining plasticity and regeneration capacity with the potential to tumorigenesis. Enrichment of KRAS copy number gain is a feature of basal subtypes (78). The KrasG12D gene dosage was positively correlated with mesenchymal cell differentiation genes. Overexpression of KrasG12D induces an epithelial-mesenchymal transition (EMT) signature and bidirectional regulation of CAF-related genes, indicating a reprogramming power of tumour cells towards basal subtypes and associated TME remodelling (79).

Epigenetic modification

Pre-neoplastic cells display accessible chromatin near oncogenes, available for later epigenetic manipulations which further drives tumour cells to either benign or malignant fates (80). Burdziak et al. found that KRAS-mutant PDAC had increased epigenetic plasticity, showing hyper-accessibility in receptor-ligand gene loci correlated with cell-cell communication and inflammation like IL-33 and IL-4 feedback signalling loops (80). Tamagawa et al. detected the trimethylation of H3K27, a genomic locus associated with endodermal development, led to ductal-to-squamous transition. They also found that KDM6A, a lysine-specific histone demethylase, which is recurrently mutant in squamous subtype and played an important role in epigenetic remodelling of PDAC cells towards squamous feature. The hyper-accessible genomic regions of KDM6A knock-off organoids were enriched with TP63 motifs, paralleled with de-enrichment of ductal/endoderm-related transcriptional factors, together with H3K27 trimethylation. This confirmed the phenotype reprogramming function of epigenetic modulation (81).

Transcriptional transition

Khoshchehreh et al. found that PDAC acquire plasticity by overexpression of the “Yamanaka factors” oct4, Sox2, klf4 and c-Myc, which induced cell pluripotency (82). Reprogramming oct4 in vitro and vivo reduced the tumorigenicity of PDAC cells, which showed potential as therapeutic target (83).

The surrogate epithelial transcription factors GATA6 and HNF4A were essential for endoderm and pancreas development. The loss of GATA6 and HNF1A/HNF4A led to a transition from Classical to squamous subtypes with increased drug tolerance (84,85).

Our previous study also found a Classical-to-Basal transition in the transcriptional profile of PDAC patients after chemotherapy. The population of GATA6 dominant Classical tumour cells decreased while KRT17 dominant Basal tumour cells persisted after chemotherapy with the present of CYP3A5, an irinotecan modulator which assisted tumour cells survive after chemotherapy (19). This reveals the tumour plasticity modulation in transcription level as a response to treatment.

Extrinsic plasticity drivers

TME and immune context

The TME influence in PDAC plasticity is observed in PDAC-derived organoids, since Basal tumour cells switch to a classical subtype after several passages, indicating external environmental redirected tumour plasticity. Moreover, the basal subtype is retained in nutritional deficient culture medium, mimicking an infertile TME. In reduced organoid medium, both Classical and Basal features were maintained, while in full organoid culture medium, referring to a nutritive TME, tumour cells lost Basal features with increased sensitivity to SN-38 and paclitaxel. These findings show that the TME has a predominant potential to manipulate PDAC cell plasticity (86).

It is also shown that tumours with higher malignant Classical or intermediate features harbour greater TME diversity. The intermediate cells are positively correlated with a T cell population, especially INFG-secreting CD8⁺ T cells. On the contrary, Basal tumours nested in more homogeneous TME, with TMA dominant but CD4⁺CD8⁺ T cell depletion (86). This suggests the potential power of the TIME towards tumour plasticity. The ECM also influences tumour plasticity. Increased deposition of ECM and thick collagen fibres activate the mechanoresponsive signatures in transitional/Basal/squamous tumour cells and eventually trigger a feedforward loop to further increase stiffness. This ECM-tumour cell interaction pushed PDAC towards a basal subtype embraced with stiff CAFs (87).

Metabolic mediation

PDAC cells dwell in a relatively infertile TME with hypoxia and glycogen deficiency with a time-related culture trajectory to squamous transition under a hypoxia state. Basal like tumour cells can better endure deprived environment and survive than Classical tumour cells. Under oxygen and Wnt-deficient environment, squamous subtype cells tend to thrive rather than Classical cells, forming a dense solid tumour instead of granular ductal shapes. The transcriptional profile gradually changes from endodermal signatures and canonical Wnt targets towards an intermediate state, with an ill-defined pattern including increased inflammation and stress-related genes, terminating into squamous/Basal signatures in late culture (81). Brunton et al. also found that the squamous tumour cells were enriched for fatty acid biosynthesis, while squamous tumour cells were enriched for glycolysis and strongly associated with increased glucose uptake and lactate production, suggesting a better glycolysis utilization than Classical tumour cells. Therefore, the deserted TME forced tumour cells develop towards squamous subtype which perform a better way of energy uptake and more likelihood to survive (85).

In addition, the hypoxic and acid TME also influences the immune profile Lactate functions as a key signalling molecule by an epigenetic modification called lysine lactylation (Kla), which in PDAC is associated with increased macrophage and Treg infiltration, and reduced CD8⁺ T cells (88). The lactate importer SLC16A1 of Tregs was found to be activated with enhanced lactate metabolization function and T cell suppression, whilst lactate induced arginase-positive macrophages were propagated to form an anti-tumour TIME (10,88). Another SLC family member SLC4A4 acted oppositely as a bicarbonate importer and showed correlation with an increased INFG-secreted effector CD8⁺ T cell infiltration as well as CD8⁺/Treg cell ratio (89). This bidirectional regulation of immune cells in response to severe TME of PDAC shows targeting potential of TME specific immunotherapy.

Throughout the heterogeneity and plasticity, PDAC tumour cells show similarity in clustered transcriptional profiles and morphological patterns can be categorised within one subgroup. PDAC neoplastic phenotypes may be classified according to transcriptional programmes into classical, basal and hybrid subtypes:

• Classical subtype: this has highly expressed epithelial markers like GATA6 and HNF1A/4A with relatively better patient outcomes. Patients with classical subtype tumours have been found to be more sensitive to erlotinib and to mFOLFIRINOX (90,91). Classical subtype tumours have a relatively immune-rich stroma and may respond better to immunotherapy (92).
• Basal subtype: this has a more mesenchymal signal expression like KRT5/17, upregulation of mutational KRAS, MYC amplification, and increases expression of PD-L1, VEGF, TGF-β, and IDO1, with a relatively worse prognosis, and sensitivity to gemcitabine-based chemotherapy (93). Basal subtype tumours have an ECM-rich stroma and are also more resistant to immunotherapy (92).
• Hybrid subtype: these tumours have mixed features of classical and basal subtypes (GATA6^Hi-KRT17^Hi) with high plasticity, and relative resistant to mFOLFIRINOX but not gemcitabine-based treatment (19).

Most PDAC tumours present a mixture of all above subtypes with the possibility of interconversion and contribute to distinct TIMEs (94-96). Classical subtypes are gland-forming tumours composed of a highly abundant ECM and CAF-rich tumour stroma, whereas basal-like subtypes are mesenchymal tumours with higher tumour cell cellularity and less ECM deposition indicating that PDAC subtypes are capable of shaping their distinct TMEs, but the underlying mechanisms are poorly understood (42).

PDAC cells exhibit considerable plasticity and continued passaging of patient derived organoids may drive more Classical phenotypes—which may differ from the original sampled tumour tissue (86,97-99). Xenotransplantation of patient derived organoids may also shift PDAC subtype which may ultimately educate the TIME (100). Raghavan et al. observed a strong culture-specific bias in PDAC-derived organoids according to different culture media in vitro. After several passages, basal and intermediate tumour cells exhibited low propagation rate, while the Classical tumour cells took the majority of the population with the continuation of the culture time (86). The transcriptional profile indicated corresponding changes with a striking loss of Basal genes such as S100A2 and KRT17, and a minimum content of intermediate Classical-Basal co-expression programs. Meanwhile, Classical genes like GATA6 and TFF1 remained largely unchanged throughout all passages. Although the KRAS copy number gain (a hallmark of basal subtype) maintained in ex vitro sample amplification, the tumour cell state shifted from basal to classical subtype in organoid culture environment, suggesting KRAS amplification only is not enough to maintain basal subtype in response to TME change. Besides, multiple gene sets associated with tumour-TME interaction and cell-cell communication (INFγ signalling, ECM organization, antigen presentation, EMT, cytokine signalling, cell adhesion and cell junction organization) were de-enriched in organoid culture environment. However, after changing the complete organoid media back to minima l media, most basal-related genes were re-enriched, suggesting the tumour plasticity manipulation by the TME (86).

Raghavan et al. then found out that non-genetic factors notable the TME for instance can strongly influence cell state and tumour response to SN-38 and paclitaxel, suggesting that tumour subtype plasticity was largely affected by TME and would further influence chemotherapy response (86). Chemotherapy can also influence tumour plasticity and the immune state (19). The basal neoplastic subtypes are significantly associated with immunosuppressed TIME with elevated numbers of macrophages and Tregs (20,101). Chemotherapy may switch pancreatic cancers from Classical dominant to a basal/Hybrid dominant type, with selective omission of GATA6^Hi-KRT17^Hi-CYP3A⁺ persister cells that cause irinotecan-based resistance (FOLFIRINOX) (19). KRT17^hi/CXCL8⁺ hybrid cells can also induce myeloid cell infiltration, switching the TME into a more immunosuppressive niche (102). Werba et al. found that chemotherapy could decrease TIGIT-CD8⁺ cell interactions and thus impede immune responses (103).

These studies indicate the complex feedback loops involved in tumour plasticity-TME-chemotherapy interactions. The next phase of research should be focused on untangling the complex interactions that drive the adaptation into a more immuno-/chemo-sensitive state: (I) transitioning tumour cells from a basal subtype to a relatively benign classical subtype; (II) educating the TIME from an immune-cold state into a hot one with more effector/cytotoxic T cell infiltration; and (III) blocking tumour cells from becoming chemoresistant by combining chemotherapy with more effective immunotherapy schemes.

Falcomatà et al. identified a synergistic interaction between the MEK inhibitor trametinib and the multi-kinase inhibitor nintedanib, which targets KRAS-directed oncogenic signalling in mesenchymal PDAC, which promoted intra-tumoral infiltration of cytotoxic and effector T cells, with to PD-L1 ICIs (104). Faraoni et al. identified that the ecto-5-nucleotidase a cell-surface enzyme (Nt5e/CD73), required for extracellular adenosine generation, was highly expressed in activated CD8⁺ T cells in PDAC tumours and was associated with adenosine-mediated immunosuppression along with Adora2b adenosine receptor expression (105). High NT5E expression is also found in squamous and basal subtypes, with known immunosuppressive and poor prognostic characteristics, altogether indicating that CD73 and Adora2b are both potential immunotherapeutic targets (105,106). The basal-like subtype was sustained via BRD4 (bromodomain-containing protein 4) mediated cJUN/AP1 expression, which induces CCL2 to recruit TNF-α-secreting macrophages forcing the Classical neoplastic cells into an aggressive phenotypic state via lineage reprogramming (77,86,107). Macrophages targeted using AZD7507, a potent selective inhibitor of colony-stimulating factor-1 receptor (CSF1R) caused tumour shrinkage, enhanced T cell activation, independent of PD-1 inhibition and with loss of TAMs the squamous gene expression programs were downregulated, with marked activation of ADEX and immunogenic gene programs (101). Clinical trials using TNF-α inhibitors and directed against macrophages however have so far not been positive in PDAC.

TME heterogeneity

The TME plays a universally essential role in malignant progression across different cancer types. For decades, the interaction between tumour cells and their milieu were well studied, turns out the TME support oncogenesis by angiogenesis, EMT, further encourage the invasion, migration and metastasis by losing cell-cell adhesion, and meantime adjust the immune system into a deficient one in favour of tumour evasion (108,109). PDAC is notorious for not only high transcriptional heterogeneity, but high TME heterogeneity as well. However, the complicated tumour-TME interactions in pancreatic cancer reveal poorly understood. In 2021, Grünwald et al. found the regional histological patterns which recurred across the PDAC patients despite of tumour subtypes, TNM stages and sites (20). The research classified the TME into three subtypes:

• Deserted TME: formed by thin, spindle-shaped fibroblasts regarded as in quiescence state, loose matured fibres with ECM-enriched, humoral immunity pathways according to transcriptional profiling, and high collagen content and CD20⁺ B cells according to immunohistochemistry (IHC) staining.
• Reactive TME: formed by plump fibroblasts with enlarged nuclei, often regarded as in a reactive state, with few acellular components and enriched inflammatory infiltration. Gene Set Enrichment Analysis (GSEA) showed increased gene enrichment in cellular stress response, growth factors with CAF-activating and immunomodulating functions, cytokines and cellular immunity, suggesting a more robust vitality in tumour interference. IHC showed significantly more T cell, macrophages, endothelial cells and CAFs infiltration, presenting a controversially T cell-inflammatory and immunosuppressive co-existing situation.
• Intermediate TME: a neutral state between deserted and reactive TME which occurs more frequently in PDAC samples.

Different TME subtypes co-exist within the same tumour, nevertheless, in a spatially confined manner. The basal subtype has a relatively homogeneous TME with CD8⁺ CD4⁺ T cell depletion and polarization from M1 to M2 TAMs, while the classical subtype has a more heterogeneous TME (86). KRAS mutations occur more frequently in the basal subtype and that MRTX1133, a non-covalent molecular inhibitor of the KRAS G12D mutant protein, can switch the TME by increasing effector T cell infiltration, decrease myeloid infiltration and reprogramme CAFs (110). The TME also presents the same convertible features as tumour subtypes after chemotherapy. Dias Costa et al. found that FOLFIRINOX treatment in patients with resectable PDAC may cause immune cell infiltration and alter the TME into an anti-tumorigenic state (111).

Certainly, more studies are required to decode the complex relationship between tumour and TMEs subtypes. The identification of the malignant nutritive/favoured sub-TME may serve to switch tumour cells into a moderate subtype and provide a basis for clinical treatment selection. Grünwald had already identified the deserted sub-TME as immune cell deficient and chemoprotective, which may assist to some extent the development of personal precise immune/chemotherapy.

There immune cell distribution also shows certain patterns and favours across PDAC TME. Immune cells are significantly more prevalent in tumour adjacent normal regions compared with healthy normal pancreas, suggesting a high inflammation state of PDAC. The adjacent normal region had higher prevalence of CD4⁺ T cells/B cells/DCs enriched tertile lymphoid structures (TLSs), which are positively correlated with prognosis and immunotherapy response. Immune cell density in tumour and tumour adjacent stroma (TAS) exhibits interpatient heterogeneity. CD8⁺ T cells, Th0 cells, Th2 cells, CD20⁺ B cells and NKT cells located more in TAS and TLSs rather than tumour regions, while CD4⁺ T memory cells, INFG-secreted CD4⁺ T cells and Tregs were more abundant in tumour regions throughout different samples. Patients with longer outcomes had a higher CD8/CD68 ratio in tumour regions, referring the negative prognostic role of TAM (112).

Moreover, immune cells seemed to exhibit a diverse distribution between Basal and Classical tumour cells. C1Q macrophages and CD4⁺CD52⁺ T cells showed up more in the vicinity of Basal cells but were excluded from Classical tumour sites (107). Tu et al. proved that TNF-α mediated crosstalk between tumour cells and macrophages assisted basal subtype maintenance and Classical-to-Basal transition (107). The single cell-sequencing data supported above histological findings that classical malignant program was weakly correlated with macrophage and neutrophil prevalence despite in the same immune community. However, unlike the histological landscape, transcriptional profile of tumour cells stratified as neural-like progenitor and neuroendocrine-like program were strongly associated with CD8⁺ T cells and/or cDC2s, the classical subtype was correlated with CD4⁺ T cells, while the squamous/Basal/mesenchymal cells were embraced with depleted CD8⁺ T cells, B cells and cDC1s (113). Despite the differences among immune landscape, the majority of the immune cells were universally dysregulated and dysfunctional, symbolized with increased exhaustion markers such as LAG3, TIGIT, TIM3 and CTLA-4 and decreased cytotoxicity (21,42).

Cancer associated fibroblasts

CAFs are fundamental to the development and control of PDAC tumours, a better understanding of which could inform improved therapeutic strategies against pancreatic cancer. There are significant limitations in our understanding of the origin of CAFs and the full range of heterogeneity of function, but amongst these we can include ECM remodelling, immune and tumour cells crosstalk, the secretion of growth and other soluble factors and the regulation of metabolic function (114,115). A variety of markers can be utilised to identify activated fibroblasts including α-SMA, desmin, fibroblast activation protein (FAP), fibroblast-specific protein 1 (FSP1), platelet-derived growth factor receptor (PDGFR), podoplanin, and vimentin. The role of CAFs is context-dependent which can be either be tumour promoting or restraining by influencing tumour growth, immunosuppression and tumour cell dissemination (59b). There are three main CAF types, myofibroblast (myCAF), inflammatory (iCAF), and antigen presenting (apCAF) (116,117). The resident pancreatic stellate cells (PSCs) of the healthy pancreas are thought to give rise to CAFs being activated during tumorigenesis, leading to proliferation and formation of a dense fibroblast compartment (115). PSC-derived CAFs are estimated to comprise less than 10% of all CAFs which may promote tumour growth. Most myCAFs originate from Gli1⁺ CAFs, and may comprise up to 30–50% of the CAFs in pancreatic cancer, with important roles in matrix depositing and immune response. Mesothelial cells form apCAFs by downregulating mesothelial features and gaining fibroblastic features, induced by the ligand interleukin 1 (IL-1) and TGF-β (118). The origin of iCAF is as yet unknown. Altogether in PDAC multiple CAFs have been characterised including are four tumour-restraining CAF subpopulations (myCAFs and Gli1⁺ CAFS etc.), ten tumour-promoting CAF subpopulations (iCAFs, apCAFs, and CD105⁺, metabolic, and FAP⁺/CXL12⁺ CAFs etc.) and two CAFs of unknown function (csCAFs and HOB6⁺ CAFs) (119) (Figure 3).

• myCAFs: these are αSMA^HiIL-6^Low expressing myofibroblasts induced by the malignant cell-derived ligand TGF-β. They are located in the peri-glandular region, associated with collagen deposition, activated by the TGF-β/SMAD2/3 pathway and have tumour supportive, chemoprotective with ECM and matrix metalloprotein enrichment and immunosuppressive roles (120,121). In this context TGF-β induces AREG (amphiregulin) expression in myCAFs, triggering an autocrine EGFR/ERBB2 response favouring a myCAF relative to an iCAF phenotype. Thus EGFR/ERBB2 inhibition would preferentially target myCAFs over iCAFs (118).
• iCAF: these are αSMA^LowIL-6^Hi expressing fibroblasts induced by the malignant cell-derived ligand IL-1. Inflammatory infiltration is associate with upregulation of IL-8/11, CXCL1/2, CCL2, CXCR12 and activation of a TNF response, TNF/NK-κB, IL2/STAT5, IL1/JAK/STAT3 pathways, located in distant desmoplasia area. iCAFs also play tumour promoting (IL-6, LIF) and immunosuppressive (IL-6, CXCL9/12, TGF-β, GAS6) role (122,123).
• apCAFs: these are CD74^hiMHC-II^hi (HLA-DRA/DPA1/DQA1) antigen presenting fibroblasts, with increased active CD4⁺ T cell population, STAT1, showing potentials as an immune-modulator through IFN-γ/STAT1/MHC-II pathway. apCAFs directly ligate and induce naive CD4⁺ T cells into regulatory T cells (Tregs) in an antigen-specific manner (119).

Figure 3 Spatial patterns of CAFs in PDAC. FAP⁺ and vimentin⁺ CAFs show regional distribution. Vimentin⁺ CAFs seem to locate in reactive TME, while FAP⁺ CAFs locate in deserted TME. FAP, fibroblast activation protein; H&E, hematoxylin and eosin; sub-TME, sub-tumour microenvironment; CAFs, cancer associated fibroblasts; PDAC, pancreatic ductal adenocarcinoma; TME, tumour microenvironment.

The interactions of CAFs with immune cells may point to therapeutic interventions. iCAFs and FAP⁺ CAFs polarize MDSCs, which inhibit effector T cell functions. Treg cells secret TGFβ which further activate myCAFs. This CAF subtype reversion may synergistically assist immunotherapy in pancreatic cancer. Moreover, CAFs derived from different sub-TME, have coordinated phenotypes with consistent behaviour patterns. The growth pattern of the CAFs in 2D culture recapitulate the characteristic histology of the sub-TMEs from which they originated: reactive sub-TME CAFs are more motile, with enriched immune-related (IL-6/JAK/STAT, interferons, TNF-α) and EMT genes, while deserted sub-TME CAFs grow faster with enrichment of growth-related and G2M-S phase genes. The deserted-intermediate-reactive transition profile shows a clear decrease of pluripotency markers (NT5E, CD44, KIT) and an increase of CAF activation/regulatory markers (αSMA, FAPα, S100A4) and inflammatory markers (IL-6, IL-1B) (20). A study based on a co-culture model of PDAC tumour cells and CAFs revealed a significant tumour cell gene shift towards proliferation and EMT, while the CAFs shifted towards the iCAF subtype with upregulated proliferation-related genes and increased interferon response genes (124). Picard and colleagues found that IL-17A-producing CD8⁺ T cells could induce iCAF and further increased tumour cell growth in vivo (125). HIFα-induced hypoxia promotes tumour growth and potentiates the iCAF TME through PDAC cell-derived cytokines (126).

A therapeutic advantage could be gained to switch to a more favourable TME by polarize the CAF phenotypes. A combination of IL-6 and PD-L1 antibody blockade showed promising tumour suppressive effect in mice model (127). A corresponding phase IB/II clinical trial og siltuximab a monoclonal antibody that inhibits IL-6 receptor binding and spartalizumab monoclonal antibody directed against the human programmed death-1 (PD-1) receptor in patients with metastatic pancreatic cancer has just completed without results posted (NCT04191421).

Conclusions

After a most unpromising start targeted personalised immunotherapies are beginning to reveal considerable impact and many more trials are now recruiting and even more are being initiated. There is a dynamic interaction in the tumour, between the PDAC cells, stromal cells, immune cells, and the subtype-TMEs. Moreover, Halbrook et al. have summarised how the metabolic adaptations that present therapeutic opportunities in pancreatic cancer are being addressed (12). The nutrient recycling and scavenging crosstalk between tumour cells, stromal cells and TMAs may unveil potential targets to tip the balance and break the alliance (12). Advances in single cell sequencing and spatially resolved imaging modalities are being used to uncover new targets for therapeutic intervention and are transforming our understanding of this difficult to treat disease (128). Di Chiaro et al. using single cell molecular deconvolution of histologically defined regions of interest identified three distinct functional morphotypes each associated with a characteristic ECM: (I) a glandular morphotypes with typical ductular structures; (II) a transitional variant with abortive ductular features and an endodermal-myofibroblast transcriptome; and (III) a poorly differentiated morphotype with absent basement membrane and ductular characteristics, and a neuronal gene expression pattern indicative of abortive lineage priming unable to undergo endodermal differentiation (87). There was evidence of dynamic plasticity and interplay between these morphotypes, governed by ECM composition and associated with local neuronal invasion (87). With the deeper understanding of the tumour plasticity and TME heterogeneity, we have identified tumour-TME interaction as a dynamic process. In exploring this transitioning interaction, we may find pointers as to how to turn the cold immune state into a hot therapeutically susceptible one. The combination of immunotherapy and chemotherapy will potentially benefit PDAC patients by manipulating tumour-TME plasticity.

Acknowledgments

Funding: None.

Footnote

Peer Review File: Available at https://cco.amegroups.com/article/view/10.21037/cco-24-72/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://cco.amegroups.com/article/view/10.21037/cco-24-72/coif). J.N. serves as an unpaid editorial board member of Chinese Clinical Oncology from April 2024 to March 2026; and he received grants or contracts from Stiftung Deutsche Krebshilfe, Bundesministerium für Bildung und Forschung, Heidelberger Stiftung Chirurgie, and Dietmar Hopp Stiftung, and participation in BioNTech-BNT321. C.S. has served on advisory boards or lectured for Astra Zeneca, Bayer, BMS, Incyte, MSD, Roche, Servier, and Taiho. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Giaquinto AN, Jemal A. Cancer statistics, 2024. CA Cancer J Clin 2024;74:12-49. [Crossref] [PubMed]
2. Springfeld C, Ferrone CR, Katz MHG, et al. Neoadjuvant therapy for pancreatic cancer. Nat Rev Clin Oncol 2023;20:318-37. [Crossref] [PubMed]
3. Hu ZI, O'Reilly EM. Therapeutic developments in pancreatic cancer. Nat Rev Gastroenterol Hepatol 2024;21:7-24. [Crossref] [PubMed]
4. Neoptolemos JP, Hu K, Bailey P, et al. Personalized treatment in localized pancreatic cancer. European Surgery 2024;56:93-109. [Crossref]
5. Strobel O, Neoptolemos J, Jäger D, et al. Optimizing the outcomes of pancreatic cancer surgery. Nat Rev Clin Oncol 2019;16:11-26. [Crossref] [PubMed]
6. Neoptolemos JP, Kleeff J, Michl P, et al. Therapeutic developments in pancreatic cancer: current and future perspectives. Nat Rev Gastroenterol Hepatol 2018;15:333-48. [Crossref] [PubMed]
7. Long GV, Menzies AM, Scolyer RA. Neoadjuvant Checkpoint Immunotherapy and Melanoma: The Time Is Now. J Clin Oncol 2023;41:3236-48. [Crossref] [PubMed]
8. Yang F, Xiao Y, Ding JH, et al. Ferroptosis heterogeneity in triple-negative breast cancer reveals an innovative immunotherapy combination strategy. Cell Metab 2023;35:84-100.e8. [Crossref] [PubMed]
9. Bockorny B, Grossman JE, Hidalgo M. Facts and Hopes in Immunotherapy of Pancreatic Cancer. Clin Cancer Res 2022;28:4606-17. [Crossref] [PubMed]
10. Bailey P, Zhou X, An J, et al. Refining the Treatment of Pancreatic Cancer From Big Data to Improved Individual Survival. Function (Oxf) 2023;4:zqad011. [Crossref] [PubMed]
11. Ullman NA, Burchard PR, Dunne RF, et al. Immunologic Strategies in Pancreatic Cancer: Making Cold Tumors Hot. J Clin Oncol 2022;40:2789-805. [Crossref] [PubMed]
12. Halbrook CJ, Lyssiotis CA, Pasca di Magliano M, et al. Pancreatic cancer: Advances and challenges. Cell 2023;186:1729-54. [Crossref] [PubMed]
13. Seo YD, Jiang X, Sullivan KM, et al. Mobilization of CD8(+) T Cells via CXCR4 Blockade Facilitates PD-1 Checkpoint Therapy in Human Pancreatic Cancer. Clin Cancer Res 2019;25:3934-45. [Crossref] [PubMed]
14. Kinker GS, Vitiello GAF, Diniz AB, et al. Mature tertiary lymphoid structures are key niches of tumour-specific immune responses in pancreatic ductal adenocarcinomas. Gut 2023;72:1927-41. [Crossref] [PubMed]
15. Zou X, Guan C, Gao J, et al. Tertiary lymphoid structures in pancreatic cancer: a new target for immunotherapy. Front Immunol 2023;14:1222719. [Crossref] [PubMed]
16. Delvecchio FR, Fincham REA, Spear S, et al. Pancreatic Cancer Chemotherapy Is Potentiated by Induction of Tertiary Lymphoid Structures in Mice. Cell Mol Gastroenterol Hepatol 2021;12:1543-65. [Crossref] [PubMed]
17. Galon J, Bruni D. Tumor Immunology and Tumor Evolution: Intertwined Histories. Immunity 2020;52:55-81. [Crossref] [PubMed]
18. Menezes S, Okail MH, Jalil SMA, et al. Cancer-associated fibroblasts in pancreatic cancer: new subtypes, new markers, new targets. J Pathol 2022;257:526-44. [Crossref] [PubMed]
19. Zhou X, An J, Kurilov R, et al. Persister cell phenotypes contribute to poor patient outcomes after neoadjuvant chemotherapy in PDAC. Nat Cancer 2023;4:1362-81. [Crossref] [PubMed]
20. Grünwald BT, Devisme A, Andrieux G, et al. Spatially confined sub-tumor microenvironments in pancreatic cancer. Cell 2021;184:5577-5592.e18. [Crossref] [PubMed]
21. Yousuf S, Qiu M, Voith von Voithenberg L, et al. Spatially Resolved Multi-Omics Single-Cell Analyses Inform Mechanisms of Immune Dysfunction in Pancreatic Cancer. Gastroenterology 2023;165:891-908.e14. [Crossref] [PubMed]
22. Chung VM, Borazanci EH, Jameson GS, et al. A SU2C catalyst randomized phase II trial of pembrolizumab with or without paricalcitol in patients with stage IV pancreatic cancer who have been placed in best possible response. J Clin Oncol 2018;36:TPS4154. [Crossref]
23. Lum LG, Thakur A, Choi M, et al. Clinical and immune responses to anti-CD3 x anti-EGFR bispecific antibody armed activated T cells (EGFR BATs) in pancreatic cancer patients. Oncoimmunology 2020;9:1773201. [Crossref] [PubMed]
24. Hopkins AC, Yarchoan M, Durham JN, et al. T cell receptor repertoire features associated with survival in immunotherapy-treated pancreatic ductal adenocarcinoma. JCI Insight 2018;3:e122092. [Crossref] [PubMed]
25. Lutz E, Yeo CJ, Lillemoe KD, et al. A lethally irradiated allogeneic granulocyte-macrophage colony stimulating factor-secreting tumor vaccine for pancreatic adenocarcinoma. A Phase II trial of safety, efficacy, and immune activation. Ann Surg 2011;253:328-35. [Crossref] [PubMed]
26. Katz SC, Moody AE, Guha P, et al. HITM-SURE: Hepatic immunotherapy for metastases phase Ib anti-CEA CAR-T study utilizing pressure enabled drug delivery. J Immunother Cancer 2020;8:e001097. [Crossref] [PubMed]
27. Lau SP, van 't Land FR, van der Burg SH, et al. Safety and tumour-specific immunological responses of combined dendritic cell vaccination and anti-CD40 agonistic antibody treatment for patients with metastatic pancreatic cancer: protocol for a phase I, open-label, single-arm, dose-escalation study (REACtiVe-2 trial). BMJ Open 2022;12:e060431. [Crossref] [PubMed]
28. Beatty GL, Chiorean EG, Fishman MP, et al. CD40 agonists alter tumor stroma and show efficacy against pancreatic carcinoma in mice and humans. Science 2011;331:1612-6. [Crossref] [PubMed]
29. Nywening TM, Wang-Gillam A, Sanford DE, et al. Targeting tumour-associated macrophages with CCR2 inhibition in combination with FOLFIRINOX in patients with borderline resectable and locally advanced pancreatic cancer: a single-centre, open-label, dose-finding, non-randomised, phase 1b trial. Lancet Oncol 2016;17:651-62. [Crossref] [PubMed]
30. Watkins DJ, Starling N, Cunningham D, et al. The combination of a chemotherapy doublet (gemcitabine and capecitabine) with a biological doublet (bevacizumab and erlotinib) in patients with advanced pancreatic adenocarcinoma. The results of a phase I/II study. Eur J Cancer 2014;50:1422-9. [Crossref] [PubMed]
31. Grierson P, Suresh R, Tan BR, et al. A pilot study of liposomal irinotecan plus 5-FU/LV combined with paricalcitol in patients with advanced pancreatic cancer whose disease progressed on gemcitabine-based therapy. J Clin Oncol 2023;41:e16263. [Crossref]
32. Kocher HM, Basu B, Froeling FEM, et al. Phase I clinical trial repurposing all-trans retinoic acid as a stromal targeting agent for pancreatic cancer. Nat Commun 2020;11:4841. [Crossref] [PubMed]
33. Ali LR, Lenehan PJ, Cardot-Ruffino V, et al. PD-1 Blockade Induces Reactivation of Nonproductive T-Cell Responses Characterized by NF-κB Signaling in Patients with Pancreatic Cancer. Clin Cancer Res 2024;30:542-53. [Crossref] [PubMed]
34. Bengsch F, Knoblock DM, Liu A, et al. CTLA-4/CD80 pathway regulates T cell infiltration into pancreatic cancer. Cancer Immunol Immunother 2017;66:1609-17. [Crossref] [PubMed]
35. Freed-Pastor WA, Lambert LJ, Ely ZA, et al. The CD155/TIGIT axis promotes and maintains immune evasion in neoantigen-expressing pancreatic cancer. Cancer Cell 2021;39:1342-1360.e14. [Crossref] [PubMed]
36. Heiduk M, Klimova A, Reiche C, et al. TIGIT Expression Delineates T-cell Populations with Distinct Functional and Prognostic Impact in Pancreatic Cancer. Clin Cancer Res 2023;29:2638-50. [Crossref] [PubMed]
37. Digomann D, Strack J, Heiduk M, et al. VISTA Ligation Reduces Antitumor T-Cell Activity in Pancreatic Cancer. Cancers (Basel) 2023;15:2326. [Crossref] [PubMed]
38. Jacoberger-Foissac C, Cousineau I, Bareche Y, et al. CD73 Inhibits cGAS-STING and Cooperates with CD39 to Promote Pancreatic Cancer. Cancer Immunol Res 2023;11:56-71. [Crossref] [PubMed]
39. Wolf Y, Anderson AC, Kuchroo VK. TIM3 comes of age as an inhibitory receptor. Nat Rev Immunol 2020;20:173-85. [Crossref] [PubMed]
40. Gulhati P, Schalck A, Jiang S, et al. Targeting T cell checkpoints 41BB and LAG3 and myeloid cell CXCR1/CXCR2 results in antitumor immunity and durable response in pancreatic cancer. Nat Cancer 2023;4:62-80. [PubMed]
41. Bärthel S, Falcomatà C, Rad R, et al. Single-cell profiling to explore pancreatic cancer heterogeneity, plasticity and response to therapy. Nat Cancer 2023;4:454-67. [Crossref] [PubMed]
42. Steele NG, Carpenter ES, Kemp SB, et al. Multimodal Mapping of the Tumor and Peripheral Blood Immune Landscape in Human Pancreatic Cancer. Nat Cancer 2020;1:1097-112. [Crossref] [PubMed]
43. Chakrabarti S, Bucheit L, Starr JS, et al. Detection of microsatellite instability-high (MSI-H) by liquid biopsy predicts robust and durable response to immunotherapy in patients with pancreatic cancer. J Immunother Cancer 2022;10:e004485. [Crossref] [PubMed]
44. Cui Zhou D, Jayasinghe RG, Chen S, et al. Spatially restricted drivers and transitional cell populations cooperate with the microenvironment in untreated and chemo-resistant pancreatic cancer. Nat Genet 2022;54:1390-405. [Crossref] [PubMed]
45. Barriga FM, Tsanov KM, Ho YJ, et al. MACHETE identifies interferon-encompassing chromosome 9p21.3 deletions as mediators of immune evasion and metastasis. Nat Cancer 2022;3:1367-85. [Crossref] [PubMed]
46. Bear AS, Vonderheide RH, O'Hara MH. Challenges and Opportunities for Pancreatic Cancer Immunotherapy. Cancer Cell 2020;38:788-802. [Crossref] [PubMed]
47. Buckley CW, O'Reilly EM. Next-generation therapies for pancreatic cancer. Expert Rev Gastroenterol Hepatol 2024;18:55-72. [Crossref] [PubMed]
48. Rojas LA, Sethna Z, Soares KC, et al. Personalized RNA neoantigen vaccines stimulate T cells in pancreatic cancer. Nature 2023;618:144-50. [Crossref] [PubMed]
49. Betof Warner A, Hamid O, Komanduri K, et al. Expert consensus guidelines on management and best practices for tumor-infiltrating lymphocyte cell therapy. J Immunother Cancer 2024;12:e008735. [Crossref] [PubMed]
50. Amaria R, Knisely A, Vining D, et al. Efficacy and safety of autologous tumor-infiltrating lymphocytes in recurrent or refractory ovarian cancer, colorectal cancer, and pancreatic ductal adenocarcinoma. J Immunother Cancer 2024;12:e006822. [Crossref] [PubMed]
51. Baulu E, Gardet C, Chuvin N, et al. TCR-engineered T cell therapy in solid tumors: State of the art and perspectives. Sci Adv 2023;9:eadf3700. [Crossref] [PubMed]
52. Leidner R, Sanjuan Silva N, Huang H, et al. Neoantigen T-Cell Receptor Gene Therapy in Pancreatic Cancer. N Engl J Med 2022;386:2112-9. [Crossref] [PubMed]
53. Yeo D, Giardina C, Saxena P, et al. The next wave of cellular immunotherapies in pancreatic cancer. Mol Ther Oncolytics 2022;24:561-76. [Crossref] [PubMed]
54. Albelda SM. CAR T cell therapy for patients with solid tumours: key lessons to learn and unlearn. Nat Rev Clin Oncol 2024;21:47-66. [Crossref] [PubMed]
55. Zhu Y, Knolhoff BL, Meyer MA, et al. CSF1/CSF1R blockade reprograms tumor-infiltrating macrophages and improves response to T-cell checkpoint immunotherapy in pancreatic cancer models. Cancer Res 2014;74:5057-69. [Crossref] [PubMed]
56. Steele CW, Karim SA, Leach JDG, et al. CXCR2 Inhibition Profoundly Suppresses Metastases and Augments Immunotherapy in Pancreatic Ductal Adenocarcinoma. Cancer Cell 2016;29:832-45. [Crossref] [PubMed]
57. Chao T, Furth EE, Vonderheide RH. CXCR2-Dependent Accumulation of Tumor-Associated Neutrophils Regulates T-cell Immunity in Pancreatic Ductal Adenocarcinoma. Cancer Immunol Res 2016;4:968-82. [Crossref] [PubMed]
58. Mitchem JB, Brennan DJ, Knolhoff BL, et al. Targeting tumor-infiltrating macrophages decreases tumor-initiating cells, relieves immunosuppression, and improves chemotherapeutic responses. Cancer Res 2013;73:1128-41. [Crossref] [PubMed]
59. Noel M, O'Reilly EM, Wolpin BM, et al. Phase 1b study of a small molecule antagonist of human chemokine (C-C motif) receptor 2 (PF-04136309) in combination with nab-paclitaxel/gemcitabine in first-line treatment of metastatic pancreatic ductal adenocarcinoma. Invest New Drugs 2020;38:800-11. [Crossref] [PubMed]
60. Marjanovic ND, Weinberg RA, Chaffer CL. Cell plasticity and heterogeneity in cancer. Clin Chem 2013;59:168-79. [Crossref] [PubMed]
61. Patil K, Khan FB, Akhtar S, et al. The plasticity of pancreatic cancer stem cells: implications in therapeutic resistance. Cancer Metastasis Rev 2021;40:691-720. [Crossref] [PubMed]
62. Lee CJ, Dosch J, Simeone DM. Pancreatic cancer stem cells. J Clin Oncol 2008;26:2806-12. [Crossref] [PubMed]
63. Shah AN, Summy JM, Zhang J, et al. Development and characterization of gemcitabine-resistant pancreatic tumor cells. Ann Surg Oncol 2007;14:3629-37. [Crossref] [PubMed]
64. Gzil A, Zarębska I, Bursiewicz W, et al. Markers of pancreatic cancer stem cells and their clinical and therapeutic implications. Mol Biol Rep 2019;46:6629-45. [Crossref] [PubMed]
65. Galindo-Vega A, Maldonado-Lagunas V, Mitre-Aguilar IB, et al. Tumor Microenvironment Role in Pancreatic Cancer Stem Cells. Cells 2023;12:1560. [Crossref] [PubMed]
66. Bubin R, Uljanovs R, Strumfa I. Cancer Stem Cells in Pancreatic Ductal Adenocarcinoma. Int J Mol Sci 2023;24:7030. [Crossref] [PubMed]
67. Mahadiuzzaman ASM, Dain Md Opo FA, Alkarim S. Stem cell-based targeted therapy in pancreatic cancer: Current approaches and future prospects. Tissue Cell 2024;89:102449. [Crossref] [PubMed]
68. Jiang H, Hegde S, Knolhoff BL, et al. Targeting focal adhesion kinase renders pancreatic cancers responsive to checkpoint immunotherapy. Nat Med 2016;22:851-60. [Crossref] [PubMed]
69. Feig C, Jones JO, Kraman M, et al. Targeting CXCL12 from FAP-expressing carcinoma-associated fibroblasts synergizes with anti-PD-L1 immunotherapy in pancreatic cancer. Proc Natl Acad Sci U S A 2013;110:20212-7. [Crossref] [PubMed]
70. Lo A, Li CP, Buza EL, et al. Fibroblast activation protein augments progression and metastasis of pancreatic ductal adenocarcinoma. JCI Insight 2017;2:e92232. [Crossref] [PubMed]
71. Padrón LJ, Maurer DM, O'Hara MH, et al. Sotigalimab and/or nivolumab with chemotherapy in first-line metastatic pancreatic cancer: clinical and immunologic analyses from the randomized phase 2 PRINCE trial. Nat Med 2022;28:1167-77. [Crossref] [PubMed]
72. Musher BL, Rowinsky EK, Smaglo BG, et al. LOAd703, an oncolytic virus-based immunostimulatory gene therapy, combined with chemotherapy for unresectable or metastatic pancreatic cancer (LOKON001): results from arm 1 of a non-randomised, single-centre, phase 1/2 study. Lancet Oncol 2024;25:488-500. [Crossref] [PubMed]
73. Zhou X, Hu K, Bailey P, et al. Clinical Impact of Molecular Subtyping of Pancreatic Cancer. Front Cell Dev Biol 2021;9:743908. [Crossref] [PubMed]
74. Braxton AM, Kiemen AL, Grahn MP, et al. 3D genomic mapping reveals multifocality of human pancreatic precancers. Nature 2024;629:679-87. [Crossref] [PubMed]
75. Alonso-Curbelo D, Ho YJ, Burdziak C, et al. A gene-environment-induced epigenetic program initiates tumorigenesis. Nature 2021;590:642-8. [Crossref] [PubMed]
76. Quintanal-Villalonga Á, Chan JM, Yu HA, et al. Lineage plasticity in cancer: a shared pathway of therapeutic resistance. Nat Rev Clin Oncol 2020;17:360-71. [Crossref] [PubMed]
77. Bailey P, Chang DK, Nones K, et al. Genomic analyses identify molecular subtypes of pancreatic cancer. Nature 2016;531:47-52. [Crossref] [PubMed]
78. Chan-Seng-Yue M, Kim JC, Wilson GW, et al. Transcription phenotypes of pancreatic cancer are driven by genomic events during tumor evolution. Nat Genet 2020;52:231-40. [Crossref] [PubMed]
79. Mueller S, Engleitner T, Maresch R, et al. Evolutionary routes and KRAS dosage define pancreatic cancer phenotypes. Nature 2018;554:62-8. [Crossref] [PubMed]
80. Burdziak C, Alonso-Curbelo D, Walle T, et al. Epigenetic plasticity cooperates with cell-cell interactions to direct pancreatic tumorigenesis. Science 2023;380:eadd5327. [Crossref] [PubMed]
81. Tamagawa H, Fujii M, Togasaki K, et al. Wnt-deficient and hypoxic environment orchestrates squamous reprogramming of human pancreatic ductal adenocarcinoma. Nat Cell Biol 2024;26:1759-72. [Crossref] [PubMed]
82. Khoshchehreh R, Totonchi M, Carlos Ramirez J, et al. Epigenetic reprogramming of primary pancreatic cancer cells counteracts their in vivo tumourigenicity. Oncogene 2019;38:6226-39. [Crossref] [PubMed]
83. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 2006;126:663-76. [Crossref] [PubMed]
84. Kloesch B, Ionasz V, Paliwal S, et al. A GATA6-centred gene regulatory network involving HNFs and ΔNp63 controls plasticity and immune escape in pancreatic cancer. Gut 2022;71:766-77. [Crossref] [PubMed]
85. Brunton H, Caligiuri G, Cunningham R, et al. HNF4A and GATA6 Loss Reveals Therapeutically Actionable Subtypes in Pancreatic Cancer. Cell Rep 2020;31:107625. [Crossref] [PubMed]
86. Raghavan S, Winter PS, Navia AW, et al. Microenvironment drives cell state, plasticity, and drug response in pancreatic cancer. Cell 2021;184:6119-6137.e26. [Crossref] [PubMed]
87. Di Chiaro P, Nacci L, Arco F, et al. Mapping functional to morphological variation reveals the basis of regional extracellular matrix subversion and nerve invasion in pancreatic cancer. Cancer Cell 2024;42:662-681.e10. [Crossref] [PubMed]
88. Peng T, Sun F, Yang JC, et al. Novel lactylation-related signature to predict prognosis for pancreatic adenocarcinoma. World J Gastroenterol 2024;30:2575-602. [Crossref] [PubMed]
89. Cappellesso F, Orban MP, Shirgaonkar N, et al. Targeting the bicarbonate transporter SLC4A4 overcomes immunosuppression and immunotherapy resistance in pancreatic cancer. Nat Cancer 2022;3:1464-83. [Crossref] [PubMed]
90. Aung KL, Fischer SE, Denroche RE, et al. Genomics-Driven Precision Medicine for Advanced Pancreatic Cancer: Early Results from the COMPASS Trial. Clin Cancer Res 2018;24:1344-54. [Crossref] [PubMed]
91. Collisson EA, Sadanandam A, Olson P, et al. Subtypes of pancreatic ductal adenocarcinoma and their differing responses to therapy. Nat Med 2011;17:500-3. [Crossref] [PubMed]
92. O'Kane GM, Grünwald BT, Jang GH, et al. GATA6 Expression Distinguishes Classical and Basal-like Subtypes in Advanced Pancreatic Cancer. Clin Cancer Res 2020;26:4901-10. [Crossref] [PubMed]
93. Maurer C, Holmstrom SR, He J, et al. Experimental microdissection enables functional harmonisation of pancreatic cancer subtypes. Gut 2019;68:1034-43. [Crossref] [PubMed]
94. Williams HL, Dias Costa A, Zhang J, et al. Spatially Resolved Single-Cell Assessment of Pancreatic Cancer Expression Subtypes Reveals Co-expressor Phenotypes and Extensive Intratumoral Heterogeneity. Cancer Res 2023;83:441-55. [Crossref] [PubMed]
95. Lee JJ, Bernard V, Semaan A, et al. Elucidation of Tumor-Stromal Heterogeneity and the Ligand-Receptor Interactome by Single-Cell Transcriptomics in Real-world Pancreatic Cancer Biopsies. Clin Cancer Res 2021;27:5912-21. [Crossref] [PubMed]
96. Connor AA, Denroche RE, Jang GH, et al. Association of Distinct Mutational Signatures With Correlates of Increased Immune Activity in Pancreatic Ductal Adenocarcinoma. JAMA Oncol 2017;3:774-83. [Crossref] [PubMed]
97. Juiz N, Elkaoutari A, Bigonnet M, et al. Basal-like and classical cells coexist in pancreatic cancer revealed by single-cell analysis on biopsy-derived pancreatic cancer organoids from the classical subtype. FASEB J 2020;34:12214-28. [Crossref] [PubMed]
98. Tiriac H, Belleau P, Engle DD, et al. Organoid Profiling Identifies Common Responders to Chemotherapy in Pancreatic Cancer. Cancer Discov 2018;8:1112-29. [Crossref] [PubMed]
99. Boj SF, Hwang CI, Baker LA, et al. Organoid models of human and mouse ductal pancreatic cancer. Cell 2015;160:324-38. [Crossref] [PubMed]
100. Miyabayashi K, Baker LA, Deschênes A, et al. Intraductal Transplantation Models of Human Pancreatic Ductal Adenocarcinoma Reveal Progressive Transition of Molecular Subtypes. Cancer Discov 2020;10:1566-89. [Crossref] [PubMed]
101. Candido JB, Morton JP, Bailey P, et al. CSF1R(+) Macrophages Sustain Pancreatic Tumor Growth through T Cell Suppression and Maintenance of Key Gene Programs that Define the Squamous Subtype. Cell Rep 2018;23:1448-60. [Crossref] [PubMed]
102. Carpenter ES, Kadiyala P, Elhossiny AM, et al. KRT17high/CXCL8+ Tumor Cells Display Both Classical and Basal Features and Regulate Myeloid Infiltration in the Pancreatic Cancer Microenvironment. Clin Cancer Res 2024;30:2497-513. [Crossref] [PubMed]
103. Werba G, Weissinger D, Kawaler EA, et al. Single-cell RNA sequencing reveals the effects of chemotherapy on human pancreatic adenocarcinoma and its tumor microenvironment. Nat Commun 2023;14:797. [Crossref] [PubMed]
104. Falcomatà C, Bärthel S, Widholz SA, et al. Selective multi-kinase inhibition sensitizes mesenchymal pancreatic cancer to immune checkpoint blockade by remodeling the tumor microenvironment. Nat Cancer 2022;3:318-36. [Crossref] [PubMed]
105. Faraoni EY, Singh K, Chandra V, et al. CD73-Dependent Adenosine Signaling through Adora2b Drives Immunosuppression in Ductal Pancreatic Cancer. Cancer Res 2023;83:1111-27. [Crossref] [PubMed]
106. DelGiorno K. CD73 Is a Critical Immune Checkpoint in a Molecular Subtype of Pancreatic Cancer. Cancer Res 2023;83:977-8. [Crossref] [PubMed]
107. Tu M, Klein L, Espinet E, et al. TNF-α-producing macrophages determine subtype identity and prognosis via AP1 enhancer reprogramming in pancreatic cancer. Nat Cancer 2021;2:1185-203. [Crossref] [PubMed]
108. de Visser KE, Joyce JA. The evolving tumor microenvironment: From cancer initiation to metastatic outgrowth. Cancer Cell 2023;41:374-403. [Crossref] [PubMed]
109. Joyce JA, Fearon DT. T cell exclusion, immune privilege, and the tumor microenvironment. Science 2015;348:74-80. [Crossref] [PubMed]
110. Mahadevan KK, McAndrews KM, LeBleu VS, et al. KRAS(G12D) inhibition reprograms the microenvironment of early and advanced pancreatic cancer to promote FAS-mediated killing by CD8(+) T cells. Cancer Cell 2023;41:1606-1620.e8. [Crossref] [PubMed]
111. Dias Costa A, Väyrynen SA, Chawla A, et al. Neoadjuvant Chemotherapy Is Associated with Altered Immune Cell Infiltration and an Anti-Tumorigenic Microenvironment in Resected Pancreatic Cancer. Clin Cancer Res 2022;28:5167-79. [Crossref] [PubMed]
112. Liudahl SM, Betts CB, Sivagnanam S, et al. Leukocyte Heterogeneity in Pancreatic Ductal Adenocarcinoma: Phenotypic and Spatial Features Associated with Clinical Outcome. Cancer Discov 2021;11:2014-31. [Crossref] [PubMed]
113. Hwang WL, Jagadeesh KA, Guo JA, et al. Single-nucleus and spatial transcriptome profiling of pancreatic cancer identifies multicellular dynamics associated with neoadjuvant treatment. Nat Genet 2022;54:1178-91. [Crossref] [PubMed]
114. Lavie D, Ben-Shmuel A, Erez N, et al. Cancer-associated fibroblasts in the single-cell era. Nat Cancer 2022;3:793-807. [Crossref] [PubMed]
115. Sahai E, Astsaturov I, Cukierman E, et al. A framework for advancing our understanding of cancer-associated fibroblasts. Nat Rev Cancer 2020;20:174-86. [Crossref] [PubMed]
116. Öhlund D, Handly-Santana A, Biffi G, et al. Distinct populations of inflammatory fibroblasts and myofibroblasts in pancreatic cancer. J Exp Med 2017;214:579-96. [Crossref] [PubMed]
117. Elyada E, Bolisetty M, Laise P, et al. Cross-Species Single-Cell Analysis of Pancreatic Ductal Adenocarcinoma Reveals Antigen-Presenting Cancer-Associated Fibroblasts. Cancer Discov 2019;9:1102-23. [Crossref] [PubMed]
118. Huang H, Wang Z, Zhang Y, et al. Mesothelial cell-derived antigen-presenting cancer-associated fibroblasts induce expansion of regulatory T cells in pancreatic cancer. Cancer Cell 2022;40:656-673.e7. [Crossref] [PubMed]
119. Brichkina A, Polo P, Sharma SD, et al. A Quick Guide to CAF Subtypes in Pancreatic Cancer. Cancers (Basel) 2023;15:2614. [Crossref] [PubMed]
120. Mucciolo G, Araos Henríquez J, Jihad M, et al. EGFR-activated myofibroblasts promote metastasis of pancreatic cancer. Cancer Cell 2024;42:101-118.e11. [Crossref] [PubMed]
121. Monteran L, Erez N. The Dark Side of Fibroblasts: Cancer-Associated Fibroblasts as Mediators of Immunosuppression in the Tumor Microenvironment. Front Immunol 2019;10:1835. [Crossref] [PubMed]
122. Vaish U, Jain T, Are AC, et al. Cancer-Associated Fibroblasts in Pancreatic Ductal Adenocarcinoma: An Update on Heterogeneity and Therapeutic Targeting. Int J Mol Sci 2021;22:13408. [Crossref] [PubMed]
123. Zhang Y, Yan W, Collins MA, et al. Interleukin-6 is required for pancreatic cancer progression by promoting MAPK signaling activation and oxidative stress resistance. Cancer Res 2013;73:6359-74. [Crossref] [PubMed]
124. Ligorio M, Sil S, Malagon-Lopez J, et al. Stromal Microenvironment Shapes the Intratumoral Architecture of Pancreatic Cancer. Cell 2019;178:160-175.e27. [Crossref] [PubMed]
125. Picard FSR, Lutz V, Brichkina A, et al. IL-17A-producing CD8(+) T cells promote PDAC via induction of inflammatory cancer-associated fibroblasts. Gut 2023;72:1510-22. [Crossref] [PubMed]
126. Schwörer S, Cimino FV, Ros M, et al. Hypoxia Potentiates the Inflammatory Fibroblast Phenotype Promoted by Pancreatic Cancer Cell-Derived Cytokines. Cancer Res 2023;83:1596-610. [Crossref] [PubMed]
127. Mace TA, Shakya R, Pitarresi JR, et al. IL-6 and PD-L1 antibody blockade combination therapy reduces tumour progression in murine models of pancreatic cancer. Gut 2018;67:320-32. [Crossref] [PubMed]
128. Schreyer D, Neoptolemos JP, Barry ST, et al. Deconstructing Pancreatic Cancer Using Next Generation-Omic Technologies-From Discovery to Knowledge-Guided Platforms for Better Patient Management. Front Cell Dev Biol 2022;9:795735. [Crossref] [PubMed]