It has been identified that KRAS, TP53, CDKN2A, and SMAD4 are four major driver genes participating in the whole process of disease development [
2]. More than 90% of tumors harbor KRAS mutations, which are known to increase the tumor invasive ability by reprogramming pancreatic cell metabolism and promoting stromal reaction. Complex genetic and metabolic pathways were identified and utilized for treatment. Additionally, some receptors that are essential to certain pathways were studied for therapeutic use, including epidermal growth factor receptor (EGFR), human epidermal growth factor receptor 2 (HER2), and vascular endothelial growth factor receptor (VEGFR).
Jones S et al. first found 12 core signaling pathways in pancreatic cancer (PC) and determined that they were each genetically altered in 67 to 100% of the tumors by performing global genomic analysis [
3]. Their finding suggested that targeting downstream mediators or key nodal points might be more effective than targeting specific mutated genes. Bailey et al. revealed that the complex mutational landscape of PDAC comprised four subtypes—squamous, pancreatic progenitor, immunogenic, and aberrantly differentiated endocrine exocrine (ADEX). Squamous, pancreatic progenitor, and ADEX tumors were enriched in different mutations, such as TP53 and KDM6A, FOXA2/3, PDX1, and MNX1, and expressed preferentially at the early stage and with regulative genes involved in KRAS activation. By targeting certain immune modulators, specific mechanisms underlying the immunogenic subtype inferred future therapeutic development. Ten distinct molecular mechanisms concerning 32 recurrently mutated genes were also identified and included KRAS, TGF-β, WNT, Notch, ROBO/SLIT signaling, G1/S transition, SWI-SNF, chromatin modification, DNA repair, and RNA processing [
4]. Notta et al. found that this aggressive disease progressed rapidly as mitotic errors of complex rearrangement pattern occurred simultaneously rather than sequentially [
5]; the tumor microenvironment plays a key role in treatment resistance because it creates a mechanical barrier consisting of dense stroma with fibroblasts, leukocytes, hyaluronic acid (HA), cancer stem cells (CSCs), collagen, and some extracellular matrix protein, resulting in hypoxia and anti-angiogenesis, which promotes carcinogenesis, facilitates tumor progression and induces chemotherapeutic resistance.
Great effort has been made to combine classic chemotherapies and novel agents targeting the discovered mutations and pathways to improve the outcome of PDAC.
Current therapies for PDAC
The CONKO-001 trial showed that the addition of 6 months of gemcitabine as adjuvant therapy could prolong disease-free survival (DFS) and overall survival (OS) in patients with resected PC. The minimal survival benefit between the gemcitabine and observation groups could be explained by gemcitabine treatment after disease progression in the control arm (DFS, 13.4 vs 6.9 months,
P < 0.001; mOS, 22.1 vs 20.2 months,
P = 0.06) [
6]. Gemcitabine-based chemotherapy is the current standard adjuvant therapy, as an 11-year follow-up study showed it dramatically decreased 45% of the DFS rate and 24% of the mortality risk [
7]. The ESPAC-3 trial included 1088 patients after surgery to receive either bolus 5- fluorouracil (5-FU) followed by intravenous 5-FU or gemcitabine for 6 months. The median survival agreed with that in the CONKO-001 trial (23.0 vs 22.1 months), with no significant difference between the bolus of 5-FU and gemcitabine, but the gemcitabine arm showed a 10% improvement in OS [
8]. Nab-paclitaxel and FOLFIRINOX (folinic acid, fluorouracil, irinotecan, oxaliplatin) showed considerable efficacy in the metastatic setting [
9,
10]. Direct comparison of the efficacy in the adjuvant setting between nab-paclitaxel/gemcitabine and FOLFIRINOX for resected PDAC is still under clinical trials, including the APACT and PRODIGE trials (NCT01964430, NCT01526135). With the modern radiation technique updates, a remarkable promotion of survival was presented by higher dose radiotherapy plus chemotherapy in early-stage resected PDAC [
11]. Thus, it raises hope that chemoradiation therapy might extend the survival in patients with PDAC.
Although approximately 20% of early-staged PDAC patients undergo resection, the 5-year survival rate remains low. Tagged cells of the PC mouse model unexpectedly entered the bloodstream and seeded into the liver without the invasion of primary lesions, suggesting that PDAC is a systemic disease rather than localized. Xu and Liu et al. discovered that patients with CEA(+)/CA125(+)/CA19–9 ≥ 1000 U/mL showed no survival advantage benefit from pancreatectomy and postoperative elevation of CA125/CEA with normal CA19–9 being associated with a poor prognosis [
12,
13]. Luo et al. sequenced fucosyltransferase 3, pointing out that CEA and CA125 have the potential to be applied as biomarkers and should be routinely measured for Lewis-negative genotypes [
14]. Several phase I/II retrospective studies showed that neoadjuvant therapy benefited patients with high-risk factors by promoting secondary resection rates, and FOLFIRINOX presented valuable efficacy. However, because phase III randomized control and multicenter clinical trials are lacking, the optimized regimen for neoadjuvant therapy has not yet been validated [
15‐
20].
Gemcitabine-based therapy is the standard treatment for metastatic PDAC (MPC). Compared with gemcitabine, FOLFIRINOX showed a survival advantage for good-performance patients, and monotherapy S-1 presented well-tolerated, non-inferiority OS in patients with gemcitabine-refractory PDAC [
10,
21]. Nab-paclitaxel/gemcitabine demonstrated an improved response rate (23%), progression-free survival (PFS) (5.5 months), OS (8.5 months), and a better HR of 0.72 compared with other phase III studies of gemcitabine-based therapy. Although erlotinib could decrease tumor growth, prevent metastasis, and improve the anticancer effect of gemcitabine to some extent, with only a 10-day benefit, it is not routinely applied in the clinical setting because the biomarkers for a treatment response are lacking despite skin rash and p53 expression [
22]. The anti-EGFR monoclonal antibody cetuximab inhibits PC cell growth via antibody-dependent cell-mediated cytotoxicity and complement-dependent cytotoxicity, but a phase III study failed to prove a clinical benefit compared with gemcitabine (mOS, 6.3 vs 5.9 months,
P = 0.23; mPFS, 3.4 vs 3.0 months) [
23]. EGFR and the HER2 inhibitor lapatinib did not show obvious survival improvement in a single-arm phase II study (mPFS, 2.6 months,
P = 0.001; mOS, 5.2 months,
P = 0.023) [
24]. Thus, the study indicated that EGFR/HER2 expression, quantified by immunohistochemistry, can hardly identify tumorigenetic subgroups driven by EGFR/HER2, and biomarkers predicting a good response have not yet been detected. Familial PDAC harbors BRCA mutations, which reduce gene recombination and DNA damage repair ability. PARPi inhibits the DNA repair of cancer cells and enhances the efficacy of platinum drugs. Olaparib monotherapy presented a promising response to the second-line setting because it showed a response rate of 20% and a median OS of 9.8 months. PARPi is the first real PC-targeted drug, and it has the potential to be used for combination therapy [
25].
Therapies focused on the tumor microenvironment
The tumor microenvironment of PDAC is different from that of other solid tumors because it is rich in stroma but lacks an oxygen and blood supply. This unique environment results in disappointing drug efficacy. Thus, depleting stromal HA, enhancing drug delivery, anti-angiogenesis, and inhibiting metabolic reprogramming might prolong survival.
Hedgehog inhibitors
Activated pancreatic stellate cells (PSCs) are essential to neural invasion in PDAC. The overexpression of Sonic hedgehog, one of the Hedgehog (HH) signaling ligands, could activate the HH signaling pathway in PSCs, thus triggering the initiation of PDAC and resulting in desmoplastic stroma [
51,
52]. In recent years, clinical efforts have been devoted to smoothened targeting, which could regulate HH signaling. The addition of vismodegib to gemcitabine experienced a setback in improving survival in MPC [
53]. Vismodegib did not increase drug delivery efficacy or survival in metastatic colorectal cancer and ovarian cancer [
54]. A phase II study of saridegib plus gemcitabine for MPC also announced the termination of the trial because a higher rate of progressive disease appeared in the saridegib arm. Because reducing HH signaling promotes angiogenesis, the imbalance of epithelial and stromal elements might explain the clinical trial’s failure. HH inhibitors were likely to result in a strong reduction in signaling, but they were metabolized and excreted between doses; thus, the possible key consideration to elevate the efficacy might be the HH inhibitor dosage [
55]. However, there is no clinical trial concerning the combination therapy in this field. SHH and Gli1 are independent prognostic factors for resected PDAC because a lower expression of Gli1 or SHH led to longer DFS and OS [
56]. Thus, HH signaling showed promise to become a diagnostic tool for APC, such as NF-κB expression, MMP9 expression, and Gli1 expression and guide therapies in the future.
Therapies enhancing drug delivery
PDAC is characterized by dense stroma with a redundant amount of HA, which regulates cancer initiation, progression, angiogenesis, and chemotherapy resistance. A high expression of HA is an independent prognostic factor of resected PDAC; thus, therapeutics targeting HA might cross the drug delivery barrier [
57]. 4-MU is well recognized for its function to suppress HA synthesis and high safety profile, proving its anti-cancer activity in vitro and in vivo in other cancers [
58,
59]. 4-MU inhibited liver metastasis in mice inoculated with human PC cells, showing the prospect for it to become a PDAC treatment targeting hyaluronan in the clinical setting.
HA interacts with CD44 on the cell surface, regulating the invasion of PDAC. The CD44-HA interaction resulted in the abrogation of cellular events by digesting HA oligosaccharides, which could be a potential direction for targeted therapy. The downstream of CA44-HA interaction could reduce metastasis to the peritoneum induced through HA by the PI3K inhibitor wortmannin [
60].
Depleting stromal HA in PDAC could reduce the tumor pressure and vascular compression and improve drug delivery efficiency. A phase II study combined PEGPH20, which degrades HA, with nab-paclitaxel/gemcitabine for MPC, with promising results. The PFS was significantly higher among HA-high patients in the PEGPH20 plus nab-paclitaxel/gemcitabine arm (HR = 0.51, 95% CI 0.26–1.00, P = 0.048), and thromboembolic events were similar in both arms. These data suggested that PEGPH20 is a potential agent for combination therapy in the high-HA subgroup, and the results of the ongoing global phase III HALO 301 trial with co-primary endpoints are awaited (NCT01839487).
Hyaluronan also compresses the intratumoral microvasculature and creates a drug delivery barrier. Angiotensin receptor blocker (ARB) could reduce stromal collagen and hyaluronan production, downregulating the expression of TGF-β1, CCN2, and ET-1 [
61]. Preclinical studies concluded that it increased vascular perfusion, oxygen, and potentiates chemotherapy [
62]. A phase II clinical trial concerning FOLFIRINOX plus losartan and radiation with proton to confirm whether ARB could sensitize PDAC compared with classic chemotherapy alone is currently recruiting (NCT01821729).
Anti-angiogenesis
Preclinical studies have shown that vascular endothelial growth factor (VEGF) could bind to VEGFR to mediate signaling events, which increase endothelial cell proliferation and migration in PDAC. Favorable results of phase II studies did not lead to a satisfying outcome in phase III studies. A recombinant fusion protein, aflibercept, comprising VEGFR-1 and VEGFR-2, was used in a phase III randomized trial with gemcitabine for APC patients, but it was stopped for futility due to a short OS and frequent adverse events [
63]. Anti-VEGF-A monoclonal antibody bevacizumab plus gemcitabine compared with gemcitabine alone in APC did not improve OS or PFS in a phase III CALGB 80303 study (mOS, 5.8 vs 5.9 months,
P = 0.95; mPFS, 3.8 vs 2.9 months,
P = 0.07). Similarly, sorafenib, targeting Ras-dependent signaling, and angiogenic pathways, showed considerable efficacy and safety with gemcitabine in a phase I trial. However, a randomized phase III trial—the BAYPAN study—observed no improvement in OS upon the addition of sorafenib to gemcitabine in APC [
64]. A phase II study of the addition of VEGFRs, PDGFRs, and the SCFR inhibitor sunitinib to gemcitabine failed to improve PFS or OS but was associated with more toxicity [
65]. Axitinib, a selective oral inhibitor of VEGFR-1, − 2 and − 3, showed an ineffective outcome in improving the survival in APC patients in phase II/III studies [
66]. The contrast between preclinical studies and clinical trials may be because experimental models are rich in vascularity and lack desmoplastic reaction. The promising outcome of phase II studies was not obtained in phase III studies likely because the design of phase II studies was non-randomized and single armed. Studies of anti-angiogenic agents/gemcitabine-based therapies have achieved no superiority in survival. However, because 35% of PDACs are of the angiogenic phenotype, it is possible that a subgroup of patients would benefit from antiangiogenic therapies [
67]. One hundred fifty-six patients enrolled in CALGB 80303 underwent detection for histidine-rich glycoprotein and complement factor H to predict the response to bevacizumab. However, only histidine-rich glycoprotein was weakly associated with OS [
68]. Subsequently, three predictive biomarkers were discovered in the same group of patients. A low level of VEGF-D was a favorable factor to benefit from bevacizumab/gemcitabine, and Ang-2 and SDF-1 were favorable to the gemcitabine/placebo arm [
69]. The identification of survival-related biomarkers and discoveries of other involved angiogenic pathways may salvage the present dilemma.
Metabolic tumor burden evaluation mainly includes total lesion glycolysis and the metabolic tumor volume, which showed strong consistency with CA19–9 and clinical prognosis for resectable PDAC [
70]. Because the desmoplastic microenvironment results in hypoxia and jejuneness, PC cells metabolized 10 times more glucose than normal cells. Glucose deprivation leads to the development of mutations in genes in the KRAS pathways in PC cells, which later reprogram the cellular metabolism and sustain unrestricted tumor growth [
71]. To improve survival under this nutrient-deprived condition, autophagy is activated to promote the use of an inner source [
72]. Yang et al. demonstrated that chloroquine and its derivatives could effectively inhibit autophagy, leading to tumor regression in PDAC [
73]. Hydroxychloroquine (HCQ) has been studied in several clinical trials as single-agent therapy in the MPC setting and as neoadjuvant therapy plus gemcitabine/nab-paclitaxel in resectable PC or MPC (NCT01978184 and NCT01506973). However, HCQ did not achieve consistent inhibition of autophagy as a monotherapy agent, and whether it is effective in combination with classic chemotherapy remains uncertain. Liang et al. demonstrated that ADP-ribosylation factor 6 (ARF6) was a downstream target of KRAS/ERK signaling. The silencing of ARF6 reduced PC cell proliferation and attenuated the Warburg effect [
74].
PC cells tend to metabolize glucose through pyruvate in aerobic mitochondrial metabolism and glycolysis to convert pyruvate to lactate [
75]. CPI-163—a mitochondrial metabolism inhibitor—could suppress the activity of α-ketoglutarate dehydrogenase and pyruvate dehydrogenase. A phase I study of the addition of CPI-163 to mFOLFIRINOX for MPC compared with FOLFIRINOX alone revealed that the ORR was higher in the CPI-163 arm with no increase in toxicity (ORR 53.9 vs 31.8%, NCT01835041). Argininosuccinate synthase inhibitor ADI-PEG 20 demonstrated a median PFS of 6.1 months and a median OS of 11.3 months in MPC when treated with the phase II dose combined with nab-paclitaxel and gemcitabine in a phase I/Ib study. However, adverse events increased with the use of enzyme inhibitors; thus, concern should be raised focusing on the efficacy of the combination with chemotherapy and associated toxicity profile [
76].
Immunotherapy
The immunosuppressive microenvironment of PC is highly heterogeneous and complex. Macrophages, myeloid-derived suppressor cells (MDSCs), and regulatory T cells (Tregs) are the three major leukocyte subtypes in the early stages of pancreatic intraepithelial neoplasia, which results in the rise of Tregs and inactivation of effector T cells. Immunosuppressive cells also accumulate in the peripheral blood, stroma, and PC tissues, inhibiting the proliferation and response of normal effector T cells. In addition, NK cell immunity, FoxP3 + T cells and CD8 + T cells were positively correlated with the survival of patients. Regulation of these immune cells may inhibit or limit tumor growth [
77].
Immunotherapies against PDAC could be categorized as passive or active immune responses. Passive immune responses were more involved in impeding molecules that regulate cancer initiation, development, and progression. Mesothelin (MSLN) is a differentiated antigen that is overexpressed in more than 90% of PDACs. Anti-MSLN antibody SS1P binds to MSLN, resulting in the inhibition of protein synthesis and apoptosis. A phase I/II study of SS1P plus pentostatin and cyclophosphamide is being studied for PC and other MSLN-expressing malignancies (NCT01362790). Anti-MSLN monoclonal antibody MORAb-009 plus gemcitabine failed to demonstrate any clinical benefit compared with gemcitabine alone in a phase II study (mOS, 6.5 vs 6.9 months; mPFS, 3.4 vs 3.5 months; NCT00570713). Monoclonal antibodies MVT-5873 targeted at CA19–9 are under safety and tolerability study (NCT02672917). Cetuximab and matuzumab—two monoclonal antibodies that target EGFR—have demonstrated convincing preclinical results. The addition of cetuximab to gemcitabine in APC patients showed no clinical significance, while stable disease and a partial response were achieved with the addition of matuzumab to gemcitabine [
23].
Active immune responses could be interpreted as exposing the tumor antigen to stimulate tumor-specific immunity. A phase I/II study of the telomerase peptide vaccine GV1001 showed a survival benefit when combined with granulocyte-macrophage colony stimulating factor (GM-CSF) in unresectable PC [
78]. However, phase III studies of GV1001 plus classic chemotherapy failed to significantly prolong survival in unresectable PC compared with that in the chemotherapy-alone group (NCT00358566, TeloVac ISRCTN4382138). Pancreatic GVAX is composed of GM-CSF that is irritated by two allogenic pancreatic cancer cell lines, inducing T cells against PDAC antigens. A randomized phase II study was conducted using GVAX with low-dose cyclophosphamide (Cy) followed by CRS-207 to inhibit regulatory T cells and induce adaptive immunity [
79]. The results showed that MPC patients who received Cy/GVAX followed by CRS-207 had a survival advantage compared with those who received Cy/GVAX alone. This finding is supported by Lutz ER’s finding that immune-based therapies could convert non-immunogenic tumors into immunogenic tumors. Thus, vaccine therapy plus immune-modulating agents would show better results than single-vaccine therapy [
80]. Algenpantucel-L was irritated by two allogenic PC cell lines (HAPa-1, HAPa-2) and was engineered by retrovirus transduction to express galactosyltransferase. An open-label phase II trial concerning the addition of algenpantucel-L to standard adjuvant therapy for resected PDAC noted that patients benefited from 300 million cells/dose compared with 100 million cells/dose (1 year DFS, 81 vs 51%; 1 year OS, 96 vs 79%) [
81]. However, the phase III IMPRESS study comparing gemcitabine plus CRT with/without algenpantucel-L for resected PC did not achieve its primary endpoint (mOS, 27.3 vs 30.4 months; NCT01072981). The poor efficacy of immunotherapy might be associated with multiple immunosuppressive mechanisms.
Bispecific antibodies (BsAb) can redirect effector cells to tumor cells without MHC restriction [
82]. BsAb binds to therapeutics and tumor cells and blocks two different oncogenic mediators simultaneously [
83]. The clinical outcome of BsAb is more appealing in hematologic malignancies than in solid tumors. MT110 is a bispecific T cell engager with two scFvs that bind to the CSC marker EpCAM on tumor cells and CD3 on T cells [
84]. Cioffi et al. found that MT110 could eliminate CSCs of PC in vivo and in vitro [
85]. Azria et al. used PC xenografts in nude mice to successfully develop a BsAb targeting TNF-α/CEA plus TNF-α and radiotherapy to control tumor growth [
86]. A phase II study of BsAb targeting HER3/IGF-IR plus nab-paclitaxel and gemcitabine in MPC is currently ongoing (NCT02399137). To improve BsAb efficacy in the treatment of solid tumors, finding a specific target and extending the short half-life of BsAb are essential in future explorations.
Mycobacteria play an irreplaceable role in modulating immune responses by enhancing the crosstalk between innate and adaptive immunity. A phase II, randomized, open-label study of heat-killed mycobacterium obuense IMM-101 plus gemcitabine was well tolerated but not significantly improved in midian OS among APC patients (6.7 vs 5.6 months,
P = 0.074) [
87,
88]. However, OS was significantly improved for 2.6 months in the pre-defined metastatic subgroup in the IMM-101 plus gemcitabine group. A large, adequately powered, phase III study of IMM-101 is needed to retain clinical efficacy in treating APC.
Targeting immune checkpoint pathways are well received currently, and promising immune checkpoint inhibitors are being developed, such as CTLA-4/B7 and PD-1/PD-L1 [
77]. However, 26 locally advanced and MPC patients showed no response to ipilimumab (anti-CTLA-4) except for one patient who showed a delayed response in a phase II study [
89]. Pembrolizumab (anti-PD-1) demonstrated a durable response with the median PFS and OS not reached in PC deficient in mismatch repair, indicating that mutant neoantigens in mismatch repair-deficient cancers are sensitive to immune checkpoint blockade [
90]. Durvalumab (anti-PD-L1) showed anti-PC activity with a 12-week DCR of 21% and an ORR of 7% in a phase II study. Additionally, MDX1105-01 (anti-PD-L1) show no objective response in 14 APC patients [
91,
92]. The lack of infiltration of effector T cells might cause resistance to single-agent checkpoint inhibitor treatment in PDAC. Innovative combination therapies of immune checkpoint inhibitors are also being explored. A phase II study of MEDI4736 (anti-PD-L1) monotherapy or in combination with tremelimumab (anti-CTLA-4) in MPC was recently completed (NCT02558894). Phase I studies of durvalumab plus pexidartinib or nivolumab plus nab-paclitaxel/gemcitabine for MPC are ongoing (NCT02777710, NCT02309177). Under the treatment of ipilimumab and GVAX for previously treated PC, two patients showed stable disease (7 and 22 weeks, respectively) without a CA19–9 biochemical response in the ipilimumab arm and three patients showed prolonged disease stabilization (31, 71 and 81 weeks) with CA19–9 declined in the ipilimumab plus GVAX arm [
93]. A phase II study of nivolumab (anti-PD-1 antibody) and GVAX/CRS-207 combinations was launched in MPC to improve the T cell-specific response (NCT02243371). Checkpoint blockade in combination with GVAX has the potential for clinical use and should be evaluated in larger studies. Because T cells are not the only key cell population essential to activate the immune response and immune checkpoint inhibitors, macrophages, and MDSCs are also indispensable members of tumor-infiltrating immune cells. Thus, it could be concluded that a future direction would be well-designed combinatorial immunotherapy and therapy approaches targeted at multiple immune cells.
Chimeric antigen receptors (CARs) are the core component of chimeric antigen receptor T cells (CAR-Ts), which confer T cells to tumor antigens through the recognition of MHC-independent ligands. CAR-T cell therapy is designed to redirect a patient or donor’s T cells to destroy tumor cells, which are highly targeted and show long-term persistence in vivo [
94,
95]. It has achieved great success in hematologic malignancies but restrained efficacy in solid cancers because of antigen loss in tumor cells and reacting in an immune-suppressive environment that lacks exclusive antigens. Preclinical studies of EGFR-specific CAR-T cell therapy promote the antitumor activity of cytokine-induced killer cells against EGFR-positive malignancies [
96]. Because EGFR is overexpressed in 90% PDAC, EGFR-specific CAR-T cell therapy presents future potent therapeutic use. HER2 overexpression leads to cellular transformation and tumorigenesis associated with a poor prognosis. However, anti-HER2-monoclonal antibodies showed no significant survival improvement in PDAC. Investigators have developed HER2-specific CAR-T cell therapy targeting HER2-positive cancer. A phase I/II study of anti-HER2 CAR-modified T cells evaluating the cytokine storm response and adverse events in refractory PDAC are currently ongoing (NCT02713984). Some studies have suggested that MSLN is the receptor of CA125/MUC16, which is closely related to metastasis in PDAC [
97]. MSLN-specific CAR-T cells showed cytolytic activity toward MSLN-positive tumor cells but induce on-target/off-tumor toxicity, causing severe adverse events [
98]. A phase I study of meso-CART by vascular intervention in APC is currently recruiting APC, hoping to increase the antitumor effect by suitable formation of tumor-associated antigen-targeted-CAR-T cells and decrease the accumulation of CAR-T cells in normal MSCL tissue (NCT02706782) [
99]. The restricted efficacy of CAR-T cells is probably due to low CAR-T cell trafficking to tumors; thus, intravenous injection and incorporation of CARs with more effector molecules might help to improve the efficacy. Fourth-generation CARs are known to augment T cell activation and attract innate immune cells, still arousing enthusiasm for novel PDAC treatment [
100].
Future challenge: personalized therapies
In the past three decades, PDAC has made slow progress compared with other tumors. Classic treatments have remained gemcitabine and 5-FU-based chemo(radiation)therapy. Over the past 5 years, preclinical studies and clinical research have gradually shed light on the crosstalk between tumor cells and the complex microenvironment. Novel regimens have been developed to target PDAC oncogenesis, the tumor microenvironment, immunosuppression, signaling pathways, and DNA damage repair. Combinations of novel agents and classic chemo(radiation)therapy prolonged the survival to some extent, but the optimized treatment combinations still need validation from more high-quality clinical trials. It also highlights the importance of biomarkers associated with efficacy. Specific biomarkers assist in selecting patients with a good response and well tolerance are desperately needed, such as CA19–9, CEA, CA125, ACIN1, TNFRSCF10C, and diagnostic miRNA panels.
Dividing PDAC as four independent diseases delivers a message that different subtypes have distinct survival rate, treatment methods, and genetic characteristics. Precision treatment is based on the identification of different subtypes of PDAC, indicating that PDAC has heterogeneous and mutation-accumulated features. Because a subset of PC harbors complex rearrangement patterns with mitotic errors simultaneously rather than sequentially, it provides insight into the tumorigenesis and tumor progression for future mechanism exploration. Deep sequencing of KRAS found that a subgroup of KRAS wild-type was observed with elevated levels of mTOR pathway proteins, suggesting that mTOR inhibitors might benefit patients with KRAS wild-type PDAC. In addition, multiple KRAS mutations were detected in the same cancer cells, possibly implying that a second KRAS mutation occurred due to selective advantage and leading to stronger KRAS signaling. It is vital that we recognize these additional intratumoral KRAS mutations and associated therapeutic resistance in future clinical settings. There is no doubt that immunotherapy is the prospect of future treatment. Similar to targeted therapy, it is crucial that further investigations should be made to characterize the underlying mechanisms among patients with the immunogenic subtype, and clinical trials of combinatorial immunotherapies are encouraged after certain patient selection.