Pancreatic Cancer: Current Multimodality Treatment Options and the Future Impact of Molecular Biological Profiling

Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal malignancies with a devastating 5-year survival rate of approximately 8% for all stages combined. It is the seventh leading cause of global cancer-related death, and incidence is rising worldwide [1].

Due to its often-silent clinical nature in early stages, PDAC is commonly diagnosed at an advanced/metastatic stage (mPDAC). Today, chemotherapy is in the center of therapy in advanced stages due to high resistance to targeted approaches or immunotherapy. Surgery usually in combination with adjuvant systemic chemotherapy is the only curative approach. However, even when deemed radiologically resectable, the majority of tumors is likely to have already spread micrometastases [2].

Thus, most patients are judged palliative at the first diagnosis or relapse after curative intended surgery. The prognosis of mPDAC still remains dismal. Therefore, future advancements must obtain better patient selection for curative resection as well as adjuvant or perioperative treatment options in a multimodal approach. Additionally, new targeted and tailored therapies have the potential to improve patients’ outcome.

The anatomic conditions of the pancreas, in particular the close proximity to surrounding major blood vessels, account for the intricacy of pancreatic surgery. Even in nonmetastatic cases, these anatomic circumstances may impede or in some cases even preclude curative intent surgery. Therefore, criteria for resectability have been defined in various classification systems which usually split nonmetastatic PDAC into resectable, borderline resectable, and locally advanced disease on the basis of anatomic criteria [3, 4]. The most recent consensus definition of the International Association of Pancreatology adds a biological and a conditional component to the mere anatomic criteria of resectability (Table 1) and is now widely used in clinical practice [5]. The biological definition of borderline resectable PDAC comprises a CA 19-9 of >500 U/mL or regional lymph node metastases diagnosed by biopsy or PET-CT or an ECOG performance status of ≥2.

The main goal of surgery for PDAC is radical resection of the tumor with clear margins and clearance of the locoregional lymph nodes. For tumors in the pancreatic head, partial pancreatoduodenectomy with standardized lymphadenectomy, nowadays usually performed with preservation of the pylorus, is the standard resection [6]. If the tumor is located in the body or tail of the gland, a distal pancreatectomy with splenectomy with routine lymphadenectomy on the left side of the celiac and mesentericoportal axis is performed.

In general, all of the above-mentioned procedures can also be performed in a minimally invasive approach, either by laparoscopic or robotic surgery, but noninferiority in regard to oncologic long-term results has not yet been proven. Regarding short-term outcomes, a recent individual patient data meta-analysis of two RCTs on minimally invasive distal pancreatectomy demonstrated a reduced rate of delayed gastric emptying and a shorter length of hospital stay for the laparoscopic approach [7]. For minimally invasive partial pancreatoduodenectomy, the issue is discussed more controversially. The largest trial had to be terminated early due to an excess number of complication-related deaths in the laparoscopic group (10% vs. 2% in the open group) [8]. On the other hand, a recent meta-analysis of three RCTs demonstrated no differences between laparoscopic and open partial pancreatoduodenectomy in terms of morbidity and mortality despite a lower blood loss and longer duration of surgery in the laparoscopic group [9]. Various trials are currently on the way to shed more light on the topic of minimally invasive versus open surgery for PDAC including robotic surgery [10, 11].

The availability and frequent utilization of more effective neoadjuvant therapies also poses some novel challenges to surgeons. Resection and reconstruction of the mesentericoportal axis has been shown to be more effective compared to palliation [12] and can be performed with acceptable morbidity and mortality [13]. Therefore, venous resections are a regular feature in the armamentarium of surgeons nowadays. The discussion is more controversial regarding arterial resection and reconstruction in locally advanced cases. While some specialized centers reported their series with promising results [14], a recent meta-analysis demonstrated that arterial resection results in a higher risk of morbidity and mortality [15]. An exception is distal pancreatectomy with en bloc celiac axis resection (DP-CAR), which can be regarded as a more standardized procedure that can be performed with acceptable morbidity and mortality [16]. In a multimodal setting, DP-CAR results in better survival compared to upfront surgery [17]. Recently, a novel technique, the so-called arterial divestment, in which the dissection is performed in a layer between the visceral arteries and the remaining tumor tissue, has been described. Even in cases with suspected arterial involvement on initial imaging, this technique has the potential to achieve R0 or R1 resections avoiding arterial resection with its increased morbidity and mortality at the same time [18].

Another issue that is still controversially discussed is resection for oligometastatic disease in PDAC. Although some centers showed promising results for this approach, most of the studies report on small cohorts that have been recruited over long periods of time, which are prone to substantial selection bias [19, 20]. Furthermore, there is yet no uniform definition of oligometastatic disease, which makes comparison and interpretation of these studies even more difficult [21]. So, resection of oligometastatic disease should not be done outside clinical trials. Not least, it should be taken into consideration that surgical results are differing between institutes with their level of experience, and close interdisciplinary cooperation is crucial for optimum patient outcome.

The standard of care for resectable PDAC is upfront resection and adjuvant chemotherapy. Real-world data show that actually only about 60% of the resected patients receive adjuvant therapy due to, e.g., perioperative morbidity [22]. The mFOLFIRINOX protocol has demonstrated tremendous efficacy compared to the long-time standard gemcitabine with a median disease-free survival (mDFS) of 21.6 months compared to 12.8 months and a 3-year DFS rate of 39.7% compared to 21.4%. mFOLFIRINOX was superior in virtually all subgroups including T3/4 tumors, N+ tumors, R1 resection margins, and poorly or undifferentiated tumors [23]. In the phase III ESPAC-4 trial, capecitabine was added to gemcitabine and resulted in a marginally significant OS prolongation of 2.5 months to only gemcitabine (HR 0.82, p = 0.032), whereas recurrence-free survival (HR 0.86, p = 0.082) was not improved significantly and toxicity increased [24]. The adjuvant therapy with gemcitabine ± capecitabine remains an option for selected patients, e.g., with pre-existing polyneuropathy, senior adults, or ECOG > 1. The multicentric phase III ESPAC I trial comparing adjuvant chemotherapy (5-FU and folinic acid) and chemoradiation (40 Gy and 5-FU) could not show a benefit for adjuvant chemoradiation, so adjuvant chemoradiation is not standard of care in this setting [25].

As PDAC frequently harbors micrometastases at time of diagnosis, neoadjuvant chemotherapy could be an efficient treatment to prevent relapse. Preoperatively, patients often present in preserved general condition allowing an intensified dose density [26]. However, about 20% of patients experience progress during neoadjuvant treatment and even become unresectable [27]. The survival benefit arising from neoadjuvant therapy was addressed by several reviews: (i) a Markov decision analysis calculated a higher life expectancy (32.2 vs. 26.7 months) [28], (ii) a propensity score matched analysis depicted a positive effect on T stage, lymph nodes, and resection margins; OS was improved compared to surgery with adjuvant chemotherapy (HR 0.83) [29].

In the PREOPANC trial, upfront surgery with adjuvant gemcitabine (OP) was compared to a neoadjuvant radiochemotherapy approach (NA) in resectable and borderline resectable patients [30]. The R0 resection rate was significantly higher in the NA arm (63% vs. 31%), which translated into a survival benefit of 3.4 months (mOS OP 13.7 vs. NA 17.1 months; HR 0.74, p = 0.074). If patients actually underwent resection, neoadjuvant therapy significantly improved this survival benefit (mOS OP 16.8 vs. NA 29.9 months, p = 0.001). This study showed a strong signal for the beneficial effects of neoadjuvant treatment, thus not including recent intensified therapy options like FOLRIRINOX or gemcitabine/nab-paclitaxel.

In a Japanese cohort, the neoadjuvant treatment with the fluoropyrimidine derivate S-1 plus gemcitabine significantly improved mOS (36.72 vs. 26.65 months; HR 0.72, p = 0.015) in a phase II/III trial when compared to adjuvant S-1 [31]. The 2-year OS rate was improved by neoadjuvant treatment with 63.7% versus 52.5%. However, transferability to other non-Asian cohorts remains elusive.

Patients with borderline resectable PDAC should receive preoperative chemotherapy. However, inconsistent definitions of borderline resectability complicate comparability of published data.

Data supporting the efficacy of FOLFIRINOX versus gemcitabine plus nab-paclitaxel or vice versa as neoadjuvant chemotherapy are lacking. A meta-analysis showed very good efficacy and safety for FOLFIRINOX with an mOS of 24.2 months and an mPFS of 15.0 months [32]. The NEOLAP trial supports data on the efficacy of induction chemotherapy (with either 4 months of gemcitabine/nab-paclitaxel or 2 months of gemcitabine/nab-paclitaxel followed by 2 months FOLFIRINOX) with an mOS of 27.5 months for patients that underwent surgery afterward [33]. Furthermore, the duration of intensive systemic (induction) therapy in borderline resectable or locally advanced PDAC is still unclear. The ultimate goal in treating locally advanced PDAC (LAPC) is to achieve resectability. Chemotherapy approaches used are similar to metastatic disease and outlined in more detail below. After neoadjuvant therapy, resectability of LAPC cannot be reliably evaluated with current imaging modalities (CT, MRI, and PET-CT). Therefore, in most patients achieving at least disease stabilization, surgical exploration is mandatory to discriminate responders from nonresponders [34]. In retrospective studies, secondary resection rates of 30–60% have been achieved, particularly after FOLFIRINOX, with R0 resection possible in up to 90% [35].

The state-of-the-art therapy for metastatic PDAC (mPDAC) patients with an ECOG 0–2 is systemic chemotherapy since it improves both overall survival and quality of life. There are several established options depending on ECOG performance status, comorbidities, and patient preferences, outlined in Table 2. Patients with ECOG 0–1 benefit from intensified chemotherapy while ECOG ≥ 2 patients with marked comorbidities should preferably receive single-agent treatment or best supportive care [47].

In the phase III PRODIGE 4 trial, FOLFIRINOX significantly improved mOS (11.1 vs. 6.8 months; HR 0.57, p < 0.001) and mPFS (6.4 vs. 3.3 months; HR 0.47; p < 0.001) compared to gemcitabine. Deterioration of quality of life was significantly delayed in the FOLFIRINOX group [36, 48], although toxicity was markedly higher. This resulted in several protocol modifications with comparable survival benefits in a meta-analysis [49]. The PRODIGE 4 Intergroup trial only included patients up to the age of 75 years. Thus, there are no data from this prospective randomized trial for this group of patients regarding FOLFIRINOX efficacy and tolerability. Retrospective data suggest that modified FOLFIRINOX protocols have a similar toxicity but also similar efficacy when compared to younger patients [50].

The phase III MPACT trial proved the benefit of adding albumin-nanoparticle bound paclitaxel (nab-paclitaxel) to gemcitabine by significantly improving mOS (8.5 vs. 6.7 months; HR 0.72, p < 0.001) and mPFS (5.5 vs. 3.7 months; HR 0.69, p < 0.001). The combination caused a higher rate of grade 3/4 adverse events mainly neutropenia, neuropathy, and diarrhea. Gem/nab-paclitaxel is safe and efficient also in ECOG 2 patients and elderly [37].

Targeted therapies commonly failed in clinical trials in mPDAC patients. An exemption is gemcitabine plus the EGFR tyrosine kinase inhibitor erlotinib that showed a significant OS benefit (HR 0.82, p = 0.038). Thus, the difference is only meaningful in patients that develop skin rash ≥ grade 2 during the first 8 weeks of treatment (mOS 10.5 vs. 5.8 months) [38].

For patients with an ECOG ≥ 2 or relevant comorbidities, monotherapy should be considered. Gemcitabine is generally well tolerated and has demonstrated efficacy in several phase III trials with an mOS of approximately 6 months [36, 37].

Upon progress, patients with ECOG ≤ 2 should be offered a 2nd-line therapy. A limitation of 2nd-line protocols available is that most have been examined in patients receiving only gemcitabine in 1st line (see Table 2).

The phase III NAPOLI trial examined the combination of nanoliposomal irinotecan (nal-Iri) plus 5-FU/FA, nal-Iri alone, and 5-FU/FA alone in pretreated mPDAC patients. The combination of nal-Iri/5-FU/FA significantly improved OS compared to 5-FU/FA alone (6.2 vs. 4.2 months; HR 0.75; p = 0.039) [43].

The OFF regimen (5-FU/LV/oxaliplatin) improved mOS (5.9 vs. 3.3 months; HR 0.66; p = 0.010) and time to tumor progression (2.9 vs. 2 months; HR 0.68; p = 0.019) compared to 5-FU/LV alone [44]. The phase III PANCREOX trial compared a modified FOLFOX6 protocol with 5-FU/LV and showed no improvement of the primary endpoint mPFS (3.1 vs. 2.9 months; p = 0.99). OS was even worse in the mFOLFOX6 arm (6.1 vs. 9.9 months; p = 0.02) [45]. After gemcitabine/nab-paclitaxel, 5-FU/FA plus nal-Iri or OFF are therapeutic options in the 2nd-line, although prospective data are lacking.

Treatment decisions after the intensified FOLFIRINOX protocol are even more complex, as there are no prospective, randomized trials evaluating a 2nd-line treatment thereafter. In the PRODIGE 4 study, 47% of patients received a 2nd-line treatment, mainly gemcitabine (82.5%) or gemcitabine-based combinations (12.5%). Gemcitabine/nab-paclitaxel as 2nd-line treatment after FOLFIRONOX has only been examined in mostly retrospective analyses [51]. Likely, the efficacy may be high with an mOS of 8.8 months and an mPFS of 5.1 months in a small cohort study. However, toxicity cannot be neglected with high-grade 3/4 toxicities [46].

After failure of a 2nd-line treatment, there are very few data suggesting the benefit of a 3rd-line treatment. In the NAPOLI-1 trial, about 30% of patients received more than one previous chemotherapy. The combination of 5-FU/FA plus nal-Iri was efficacious also in this setting for patients being irinotecan naïve (no iri: HR 0.62; iri: HR 1.25) [43].

Since about one-third of patients with pancreatic cancer die from complications related to local disease progression [52], current research has focused on the improvement of local disease control by ablative radiation. Since different phase III trials comparing the effect of chemoradiation versus chemotherapy alone for advanced PDAC have led to mixed results [53, 54], the role of palliative radiation in PDAC has been subordinated compared to the promising results of new chemotherapy strategies. However, with the development of improved radiation techniques, e.g., stereotactic body radiation or intensity modulated radiotherapy, higher radiation dose with less adverse effects can be applicated for local tumor control [54]. Single-center cohort studies using these new radiation techniques show promising results. Reyngold et al. [55] could show that hypofractioned ablative radiation after induction chemotherapy led to a 2-year local tumor progression rate of 32.8% and a 2-year overall survival rate of 38% in 119 patients from the time of irradiation. In conclusion, the present ASTRO guidelines recommend the use of fractionated and hypofractionated radiation for symptom management and disease control in metastatic or local site for patients with advanced PDAC and highlight the future perspective of evolving radiation techniques [56].

In recent years, major advances have been made in exploring PDAC biology. Commonly found mutations in PDAC are KRAS (>90%), CDKN2A/2B (90%), p53 (70%), and SMAD4 (50%) [57]. To a certain extent, PDAC biology is determined by the interplay of these classical driver mutations with passenger mutations. In line, modern genome sequencing methods significantly expanded this knowledge and, e.g., allowed PDAC subtyping into four subtypes based on patterns of structural variation in chromosomes: (i) stable (20% of the PDAC), (ii) locally rearranged (30%), (iii) scattered (36%), and (iv) unstable (14%) [57]. The clinical utility of subtyping can, for example, be shown for the unstable subtype that includes germline and somatic mutations in DNA damage repair genes like BRCA1/2, PALB2, and ATM, sensitizing those tumors to targeted therapy approaches or platinum therapy [57].

Despite promising approaches, it has to be admitted that molecular profiling for precision cancer therapy has several limitations [58]. First, tumor sequencing results and availability of best-fitting therapeutic agents are often too late while the tumor progression has changed tumor biology or has led to impaired personal health state of the patient. Second, due to intratumoral heterogeneity, molecular profiling based on a patient-derived tumor sample may not be representative for overall tumor biology, and targeted therapy may affect a nondriver or passenger mutation. Last, data on targeted therapy and molecular predictors of therapy response are often poor, and further studies with larger cohorts are mandatory. This issue is additionally aggravated by the rarity of many specific tumor mutations that could potentially be targeted by new therapeutic drugs.

The transcriptome of PDAC is not solely determined by genetic alterations. Epigenetic regulatory circuits, such as DNA methylation, posttranslational modification of histones, and regulation by noncoding RNAs, are also disrupted in PDAC. As a result, important tumor suppressor genes are suppressed, and oncogenes are upregulated [59]. Epigenetic reprogramming may lead to tumor progression and formation of metastases [60].

Despite high-quality clinical trials that have compared different chemotherapy regimens for the treatment of metastatic PDAC, the present choice of individual chemotherapy has mainly focused on patients’ comorbidities and ECOG state (see above). With rising insight in tumor biology and mechanisms of chemotherapy resistance and toxicity, several promising approaches have tried to find molecular predictors to individualize chemotherapy. Furthermore, therapeutic agents to overcome tumor drug resistance are a promising approach to improve the effect of the established chemotherapy regimen.

Regarding the use of gemcitabine, several alterations in tumor gemcitabine metabolization have been identified to predict therapy response. In particular, the low expression of human nucleoside transporter protein hENT1 in pancreatic cancer showed to worsen prognosis in patients with gemcitabine treatment, leading to an mOS of 17.1 months compared to 26.2 months in patients with high tumor hENT1 expression in a retrospective analysis of patients in the ESPAC-3 trial (n = 434) [61].

Regarding chemotherapy toxicity, genetic deficiency of dihydropyrimidine dehydrogenase (DPD), which affects approximately 9% of European patients, has been associated with increased 5-FU toxicity and worse prognosis [62]. Therefore, molecular testing for DPD deficiency before 5-FU administration is part of actual clinical practice and included in present guidelines and EMA recommendation [62]. On the other hand, high tumor expression of DPD has been identified as a mechanism of resistance leading to worse therapy response to 5-FU-based chemotherapy regimen [63].

For oxaliplatin, alterations in tumor DNA mismatch repair have been identified to predict therapy response. In a single-center retrospective analysis, 26 PDAC patients receiving a platinum-based chemotherapy with BRCA1 (n = 5), BRCA2 (n = 17), or PALB2 (n = 4) were matched to a control group. The authors could show that the objective response rate based on radiological RECIST criteria was 58% in the study group compared to 21% in the control group, leading to a real-world PFS of 10.1 months versus 6.9 months (HR 0.43; p = 0.0046) [64]. Other mutations in DNA mismatch repair pathway like ATM/ATR, CHEK2, FANCC, FANCG, and RAD51 are estimated to have a similar effect on therapy response to platinum-based therapy but have not been specifically evaluated in clinical trials so far. Overall, PDAC patients DNA mismatch repair deficiency, which is present in approximately 16% of all PDAC patients, has been associated with a worse overall prognosis due to an aggressive tumor biology and a better therapy response to platinum-based therapy [65].

At last, organotypic cultures derived from patients’ PDAC samples (organoids) are a promising approach for individualized in vitro drug testing and prediction of therapy response in vivo. Frappart and colleagues [66] have established an organoid model harboring the core features of individual PDAC samples and have demonstrated successfully the correlation between in vitro and in vivo therapeutical response. Nevertheless, it has to be admitted that organoid-derived therapy decision is still far away from becoming part of standard medical care in PDAC therapy but is investigated in clinical trials [67].

The subtyping of PDAC allows for novel targeted treatment strategies. Although the common mutations KRAS, CDKN2A, and SMAD4 are still undruggable, first attempts to target mutant TP53 or to pharmacologically manipulate RAS function have been achieved [68, 69]. However, the KRAS^G12C inhibitor sotorasib already as the first targeted approach showed high efficacy, even though this mutation is rarely found in PDAC [70]. The focus remains also on downstream effectors that are frequently altered, including, e.g., BRAF, EGFR, MEK, and PI3K/AKT.

Progression and metastasis are dependent on the MAPK signaling cascade [71], and targeting of EGFR can early intervene here. But, in a phase III trial, adding the anti-EGFR antibody cetuximab to a gemcitabine-based chemotherapy did not improve outcome [72]. The EGFR inhibitor erlotinib remains the only agent directed against the MAPK pathway that has entered clinical practice.

To improve therapeutic outcome, multitargeting strategies are considered as simultaneous inhibition of the MAPK and PI3K/AKT signaling cascades with significantly reduced tumor progression in preclinical models [73] but missing clinical translation by now [74]. Inhibition of the mTOR pathway also yielded promising preclinical data [75], but stimulates MEK/ERK and AKT signaling via feedback mechanisms [76]. mTOR inhibition showed neither as single substance nor in combination with capecitabine an OS benefit [77].

Recently, it could be shown that about 4–7% of an unselected Caucasian population exhibit germ line mutation in the BRCA1 or -2 genes even without a clear family history [41, 78]. These tumors show a disturbed DNA homologous recombination leading to deficient repair of DNA double strand breaks. The tumors appear to be particularly sensitive to DNA crosslinking agents such as cisplatin or DNA repair inhibitors such as gemcitabine. This could be demonstrated in preclinical studies, but also in phase II trials examining the effect of gemcitabine plus cisplatin in mPDAC patients with a germline BRCA or PALB2 mutation [79].

The phase III POLO trial [41] examined maintenance treatment with the PARP inhibitor olaparib compared to placebo in patients with mPDAC and a germ line mutation in BRCA1 or -2. Patients had to have at least stable disease under at least 16 weeks of a platinum-based 1st-line therapy. Eighty-one percent of patients in the trial received FOLFIRINOX. In the overall population, mOS was about 19 months stressing the beneficial role of a platinum-based treatment in patients with germ line BRCA1/2 mutations. Olaparib maintenance treatment significantly prolonged PFS, the primary endpoint of the study, compared to placebo (7.4 months vs. 3.8 months; HR 0.53; p = 0.004). However, there was no difference in overall survival between the two treatment arms (19 months vs. 19.2 months; HR 0.83, 95% CI: 0.56–1.22; p = 0.3487) (see Table 1) [42]. Olaparib was well tolerated with anemia and fatigue as the most frequent ≥ grade 3 side effects.

Thus, patients with mPDAC should undergo testing for a germ line BRCA1/2 mutation early after diagnosis since in these patients a platinum-based treatment appears highly efficacious. At the moment, there is no direct comparison whether cisplatin-based or oxaliplatin-based regimens are preferable in this situation [80]. Olaparib maintenance treatment is an interesting option for these specific patients. In case of a proof of evidence of a BRCA1/2 germ line mutation, patients must be offered genetic counseling [41]. Additionally, mutations in other HRR genes such as ATM may also sensitize to PARP inhibitors [81, 82].

New therapeutical strategies focus on the combination of targeted therapies to overcome drug resistance by synergistic effects. For example, the use of antiangiogenetic agents might increase tumor hypoxia which benefits the impairment of HRR. According to this innovative approach, the combination of bevacizumab, cediranib, and olaparib has shown promising results for the therapy of ovarian cancer without increased toxicity [83]. Furthermore, synergistic pathway inhibitor strategies are developed to overcome drug resistance. In a preclinical murine model, Gout et al. [84] have developed a triple pathway inhibition concept for a synergistic inhibition of HHR for ATM-deficient PDAC. Based on this approach, HHR is impaired including the alternative HHR pathway which prevents the development of drug resistance against single targeted agents and might have the potential to be used even for non-HHR-deficient PDAC.

PDAC is a low immunogenic tumor, and only about 1% of tumors are mismatch repair deficient (dMMR)/microstellite unstable (MSI-H) [85]. There is little evidence for checkpoint inhibition also in MSI-H-PDAC. In a phase II basket study, 8 patients with MMR-deficient PDAC achieved an objective response rate of 62% [86]. In locally advanced or metastatic adenocarcinoma of the pancreas, nivolumab and nab-paclitaxel ± gemcitabine showed in a phase I study disease control in 12 of 17 patients [87]. In patients with advanced or metastatic noncolorectal MSI-H/dMMR cancers (27 different tumor types) and disease progression or intolerance to prior standard therapy, the KEYNOTE-158 study showed benefit for therapy with the PD-1 antibody pembrolizumab in the entire patient population. However, in the subgroup of patients with MSI-H-PDAC (n = 22), the median OS was 4 months with a median PFS of only 2.3 months. In patients responding to the treatment, the duration of response was long with 13.4 months. Thus, unfortunately, only a subgroup of patients with MSI-h/dMMR, but not the whole group, appears to benefit substantially from a checkpoint inhibitor treatment. So far there is no evidence that checkpoint inhibitors have any benefit at all in PDACs without dMMR/MSI-h status [88].

PDAC is a highly malignant tumor with aggressive biology, poor prognosis, and a rising incidence. Many advances in pancreatic research over the past 2 decades have improved our understanding of PDAC development and progression, with only single therapeutic breakthroughs like the implementation of state-of-the-art surgical resection combined with adjuvant chemotherapy. In the fight against pancreatic cancer, further studies with smart combination therapies, biologically motivated, and immunotherapeutic strategies are needed.

T.J.E. acted as a consultant or received lecture fees from Bayer, BMS, MSD, Sanofi, Merck Serono, Roche, Servier, and Ipsen outside the submitted work. One of his research projects is supported by Servier. N.S., M.G., F.J.H., and L.P. have no conflicts of interest to declare.

The authors have no funding sources to declare.

T.J.E., N.S., M.G., F.J.H., and L.P. were responsible for conception, literature research, writing, and drafting the manuscript.