Role of molecular biology in the management of pancreatic cancer

Abstract

Pancreatic ductal adenocarcinoma (PDAC) presents significant challenges in patient management due to a dismal prognosis, increasing incidence, and limited treatment options. In this regard, precision medicine, which personalizes treatments based on tumour molecular characteristics, has gained great interest. However, its widespread implementation is not fully endorsed in current recommendations. This review explores key molecular alterations in PDAC, while emphasizing differences between KRAS-mutated and KRAS-wild-type tumours. It assesses the practical application of precision medicine in clinical settings and outlines potential future directions with respect to PDAC. Actionable molecular targets are examined with the aim of enhancing our understanding of PDAC molecular biology. Insights from this analysis may contribute to a more refined and personalized approach to pancreatic cancer treatment, ultimately improving patient outcomes.

Key Words: Pancreatic cancer, Precision medicine, KRAS, Molecular profile, Targeted therapies, Liquid biopsy

Core Tip: Management of patients with pancreatic cancer (PDAC) is a real challenge due to a poor prognosis, a rising incidence and few therapeutic options. Precision medicine is an approach that seeks to personalize treatments to the specific molecular characteristics of individual tumors. Despite a growing interest in applying molecular precision medicine for PDAC management, large scale precision medicine is not endorsed so far by the last recommendations. We review the main actionable molecular alterations found in PDAC, highlighting the differences between KRAS-mutated and KRAS-wild-type tumors, as well as precision medicine in clinical practice and future directions for precision medicine in PDAC.

INTRODUCTION

Pancreatic ductal adenocarcinoma (PDAC) management is challenging due to its very aggressive nature that, combined with limited therapeutic options, leads to poor overall survival (OS)[1,2]. For patients with PDAC, there is clearly an unmet need for the development of innovative strategies.

Over the years, the cancer treatment landscape has been evolving, with the emergence of a promising approach based on molecular precision medicine[3] (Figure 1). This approach seeks to personalize treatments to the specific genetic and/or molecular characteristics of individual tumours, offering the potential to improve treatment outcomes and enhance patient survival[4]. Recently, several prospective studies have been conducted to assess the feasibility and effectiveness of precision medicine through next-generation sequencing (NGS) in patients with various solid cancers[5-10]. It is worth noting that only 10% to 25% of patients in these studies received specific therapy based on molecular profiling, and few have as yet truly benefited clinically from this approach.

Today, precision medicine is now fully integrated into the management strategy for several types of tumours, including lung cancer[11]. For PDAC however, results with molecular precision medicine are controversial and despite the growing interest, its large scale application is discouraged in the latest recommendations[12,13]. Moreover, implementation of molecular precision medicine for PDAC remains challenging, due to the vast variety of KRAS alterations and the rapidly progressive nature of the disease[14]. In this review, we will explore the evolving landscape of molecular precision medicine in PDAC, highlighting the main molecular alterations found in PDAC, along with the molecularly-matched therapies and their potential benefit, and finally the remaining challenges and future directions.

Figure 1 Principles of precision medicine compared to the traditional model of patient management.

MAIN ACTIONABLE MOLECULAR ALTERATIONS FOUND IN PDAC

Molecular profiling in PDAC reveals the presence of several key genetic alterations involved in carcinogenesis, primarily in four genes: KRAS, TP53, SMAD4, and CDKN2A[14]. With the exception of the KRAS G12C mutation, which is present in less than 1%-2% of PDAC cases[15], none of these genetic alterations as yet have validated targeted therapies. However, several molecular profiling studies have demonstrated that up to 25% (ranging from 12% to 25%) of PDACs harbour actionable molecular alterations[14,16-21] (Figure 2). "Actionability" of a particular molecular alteration is defined as strong clinical or preclinical evidence of the predictive benefit from a specific therapy (in any type of cancer), as measured by the European Society for Medical Oncology Scale for Clinical Actionability of molecular Targets (ESCAT) classification[22] with ESCAT I to III being considered as actionable.

Figure 2 Main actionable alterations according to KRAS status[32,35,61,86].

The main actionable alterations primarily involve the DNA damage response and repair pathways (BRCA1, BRCA2, PALB2, ATM), targeted by platinum agents or poly (ADP-ribose) polymerase (PARP) inhibitors. The randomized phase III Polo study demonstrated a significant improvement in progression-free survival on platinum-based chemotherapy for patients with germline BRCA1/2-mutation and metastatic PDAC receiving the PARP inhibitor Olaparib[12,23]. Consequently, guidelines now recommend the testing of PDAC patients for pathogenic germline alterations in BRCA1 and BRCA2. To date, olaparib remains the only approved molecularly-targeted therapy in PDAC.

Clinical benefits from and significant responses to biomarker-based tumour agnostic therapies have also been reported for PDAC. These include immune checkpoint inhibitors for high microsatellite instability tumours[24,25] or neurotrophic tropomyosin receptor kinase (NTRK) inhibitors for NTRK1, NTRK2, NTRK3, or ROS1 fusions[26,27]. Additionally, patients with PDAC carrying BRAFV600E mutations have been shown to benefit from treatment with a RAF-MEK targeted therapy[28] or, in the case of FGFR2 fusion, from FGFR2 inhibitors[29]. Durable responses to zenocutuzumab were also observed in cases of PDAC with NRG1 fusion[30]. However, while the anti-HER2 antibody-drug conjugate trastuzumab-deruxtecan is effective in a very large range of tumours with HER2 amplification, results are disappointing in PDAC with only 4% objective responses and 36% of tumours showing controlled progression at 12 wk[31]. The major actionable molecular alterations found in PDAC are summarized in Table 1 along with their current ESCAT classification.

Table 1 Main actionable molecular alterations in pancreatic ductal adenocarcinoma and their current European Society for Medical Oncology scale for clinical actionability of molecular target classification.

Biomarker	Targeted therapy	ESCAT classification	Ref.
BRCA1, BRCA2 germinal	Olaparib	IA	[87]
BRCA1, BRCA2 germinal/somatic	Rucaparib	IB	[86,88]
PALB2 germinal/somatic	Rucaparib	IB	[88]
NTRK fusions	Larotrectinib, entrectinib	IC	[27,89]
MSI high	Pembrolizumab	IC	[25]
RET fusions	Pralseltinib, selpercatinib	IC	[90,91]
NRG1 fusions	Zenocutuzumab	IIB	[30]
KRASG12C	Sotorasib, adagrasib	IIB	[55,56]
FGFR2 Fusions	Pemigatinib, infigratinib, futibatinib	IIIA	[92,93]
PIK3CA mutations	Alpelisib, buparlisib	IIIA	[94,95]
BRAFV600Emutation	Dabrafenib + trametinib, Binimetinib + encorafenib	IIIA	[96,97]
ALK/ROS1 fusions	Crizotinib, ceritinib, alectinib	IIIA	[98-100]
ERBB2 amplification	Trastuzumab + pertuzumab, T-DM1	IIIA	[101,102]

ESCAT: European Society for Medical Oncology Scale for Clinical Actionability of molecular Targets; MSI: Microsatellite instability.

IMPACT ON PROGNOSIS OF MOLECULAR PRECISION MEDICINE

Improved overall response rates, progression-free survival, and median OS have been observed in metastatic PDAC patients with actionable molecular alterations treated with corresponding molecularly-matched treatment, compared (though not randomized in most cases) to those not receiving targeted therapy[20,21,32,33]. However, it remains to be demonstrated whether molecular profiling could improve therapeutic decision-making in PDAC patients.

Several prospective studies have attempted to assess the feasibility and clinical benefits of molecular precision medicine in patients with various advanced solid tumours[5-8]. A wide range of actionable molecular alterations has been described for most tumour types, and are found in 25% to 90% of patients[5-8]. However, only 10% to 25% of patients in these studies received molecularly-matched treatment based on precision medicine[5-10]. In PDAC, while around 25% of tumours were found with actionable targets[20,21,34], the number of patients actually receiving a molecularly-matched treatment could be as low as 5%[21]. Additionally, two clinical trials designed to evaluate precision medicine strategies (across all solid tumour types) showed controversial results[5,6]. Nevertheless, one prospective randomized precision medicine trial conducted in patients with breast cancer did report improved progression-free survival obtained from administering targeted therapies according to genomic data but only when molecular alterations were classified as level I/II according to the ESCAT scale [(Hazard Ratio (HR) 0.41, P < 0.001][8]. This has been confirmed in the data from the precision medicine program at Gustave Roussy[35].

In their retrospective study, Pishvaian et al[21] reported the OS results of the Know Your Tumour program for pancreatic cancers and found a positive impact on OS of molecularly-matched treatments, when compared to patients with an actionable molecular alteration who did not receive the corresponding targeted therapy (2.58 years vs 1.51 years). Similarly, the median OS was better when compared to that of patients without actionable molecular alterations who received standard-of-care treatment (2.58 years vs 1.32 years, HR = 0.34, P < 0.0001). There was no significant difference in OS between patients with actionable molecular alterations who did not receive the corresponding targeted therapy and those without actionable molecular alterations who received standard-of-care treatment (HR = 0.82, P = 0.10).

A similar study conducted in a reference tertiary centre involved 28% of patients with an actionable molecular alteration, 41% of whom (only 10% of the whole cohort) received a molecularly-matched treatment[36]. While no significant improvement in survival was noted, the number of treated patients was limited. In specific subgroups (Microsatellite instability[24] or FGFR2[29]), a significant improvement in survival was however noted. Thus, while PDAC may not as yet have benefited greatly from the advent of precision medicine, efforts are ongoing to improve the prognosis of patients with actionable molecular alterations. Remaining challenges are the rapid progression of the disease along with general state degradation, the vast proportion of tumours having powerfully oncogenic and largely undruggable KRAS mutations, and the large range of different actionable molecular alterations for which no Food and Drug Administration or European Medicines Agency-approved drug yet exists.

PRECISION MEDICINE IN ROUTINE HEALTHCARE

For an optimal implementation of precision medicine in routine healthcare, several points need to be considered.

First, the increasing number of molecular alterations identified in tumours, including PDAC, amplifies the challenge of their clinical interpretation. The actionability of any given molecular alteration evolves with clinical and preclinical evidence. To facilitate the interpretation of these data, a Molecular Tumour Board (MTB) has now been organized in some hospitals[37]. This MTB brings together clinicians, geneticists, molecular pathologists, genetic counsellors, bioinformaticians, and even researchers[38]. It reviews the patient's clinical, pathological, and molecular results and makes recommendations for targeted treatment options. It also classifies alterations according to ESMO recommendations using the ESCAT classification[22] and may guide the patient toward inclusion in precision medicine clinical trials.

Second, tumour-based molecular profiling can reveal incidental germline variants that are clinically significant in up to 12% of patients[39,40]. This can raise significant challenges to physicians who may not be prepared to handle these incidental findings, which can have significant implications for the patient's relatives, especially in PDAC with the involvement of BRCA1/2. Therefore, it is essential that NGS-based tumour tests are associated with appropriate pre-test counselling to inform patients about the possibility of identifying a hereditary cancer predisposition. Several studies have shown that, even in the context of advanced cancer, the majority of patients wish to be informed about such incidental findings[39,40]. For such patients, a post-test genetic counselling should be organized to plan for confirmatory germline testing with informed consent, counsel the patient on the associated risks for themselves and their relatives, and even invite relatives for result confirmation consultations. MTB could help guide patients towards oncogenetics testing to ensure that no potential germline mutation is left unexplored.

The third point for consideration in the implementation of precision medicine is NGS of circulating tumour DNA (e.g. liquid biopsy) shown to offer several advantages over tissue-based NGS in cancer molecular profiling. Indeed, tissue-based tumour sequencing, the current gold standard for detecting actionable molecular abnormalities in patients with solid cancers[41], has several limitations including screening failures due to limited tissue availability (e.g., after a biopsy) and an inability to capture spatial and temporal tumour heterogeneity, which can compromise accurate treatment selection. Furthermore, obtaining a tumour biopsy is challenging and can lead to adverse events not encountered with blood sampling[42,43]. Liquid biopsy on the other hand is non-invasive, easily feasible and reproducible, and is representative of the patient's entire tumour molecular landscape[44,45]. It allows the detection of genomic alterations with high accuracy compared to tissue analysis[44,46], and importantly captures spatial and temporal heterogeneity, a well-established characteristic of malignant tumours. The molecular profile obtained from a tumour biopsy is limited to a single site (space) and is frozen in time. Liquid biopsy has the advantage of providing access to a "genomic pool" from multiple metastatic sites in the patient[44], allowing for more precise genotyping in metastatic patients. Indeed, tumour progression and systemic treatment regimens can promote clonal evolution and increased tumour heterogeneity, leading to discordance between archived tissue and liquid biopsy results. However, one limitation of liquid biopsy is the possibility of false negatives (i.e., not identifying an alteration of interest actually present in the tumour). Indeed, Sugimori et al[47] showed fluctuations in KRAS detection in advanced PDAC patients under treatment. More data from prospective studies on the use of liquid biopsy for molecular profiling are now needed to validate its use for non-invasive serial sampling of tumour material, which could become part of PDAC management. One promising liquid biopsy test developed by Foundation Medicine^® (Massachusetts, United States) can detect targetable alterations using NGS to study 324 cancer-associated genes, including short pathogenic variants, copy number alterations, and some rearrangements, as well as mutational burden. The development of such liquid biopsy tests will likely facilitate precision medicine in PDAC.

The fourth point for consideration is access to molecular biology technologies. A recent study assessed access to molecular biology technologies in Europe and showed profound disparities between countries, some of which thus had limited access for patients to targeted drugs[48]. While immunohistochemistry and basic genomic techniques such as fluorescence in situ hybridization, polymerase chain reaction, and microsatellite instability analysis were widely available in routine clinical practice, access to more novel and advanced technologies, including large-scale NGS panels, was highly heterogeneous in Europe and within each country. Even in high-income countries, large high-throughput sequencing panels remain largely unavailable in routine clinical practice and are limited to clinical trials, research or tertiary centres. Thus, despite the existence of targeted therapies, the search for additional biomarkers in every patient is currently not permitted in routine medical practice. With the advent of new targeted therapies requiring associated biomarkers, the diffusion of molecular profiling is essential to ensure equal access to healthcare throughout Europe.

The final point for consideration is drug access issues, which at least in part explain the low percentage of patients treated with target therapies. In France, despite European marketing authorizations and studies proving their clinical effectiveness, certain innovative drugs such as larotrectinib[48] have been issued unfavourable opinions by French authorities with regards their reimbursement. Precision medicine goes hand in hand with associated medications, so advocating for NGS for all PDAC patients would only make sense should patients be assured access to treatments. While some clinical trials undertaken in France are allowing patients to receive such treatments, the restrictive inclusion criteria of these trials effectively exclude a certain number of patients.

TOWARDS PRECISION MEDICINE FOR ALL PATIENTS WITH PDAC

In 2020, it was estimated that nearly 20 million new cancer cases occurred worldwide[49,50]. Studies have revealed that approximately one in seven human cancers carries a KRAS gene mutation, making it one of the major human oncogenes[51-54]. Targeting KRAS has been a major focus of cancer research over the past four decades, particularly since the fundamental discovery by Ostrem et al[54] of a KRAS G12C inhibitor[54], which led to the clinical validation of two targeted therapies sotorasib and adagrasib[55,56]. There is a wide diversity of KRAS gene mutations[57]. While the most frequent alteration in PDAC affects codon G12, representing 93% of cases, the amino acid substitution differs among patients[58,59]. The most common substitution is G12D (41%), followed by G12V (34%) then G12R (16%) and only 1.1% for the G12C that is actionable[58,59]. The presence of the KRAS mutation is associated with a poor response to treatment and confers an unfavourable overall prognosis and shorter survival in PDAC patients[60-62]. Its targeting therefore represents a challenging yet crucial goal in the development of effective new therapies.

In 2018, sotorasib (AMG510) became the first KRAS G12C inhibitor to enter human clinical trials and demonstrated its safety and clinical effectiveness. Hong et al[55] reported the results of a phase 1 multicentre open-label trial[55] involving sotorasib in patients with advanced solid tumours (locally advanced or metastatic) harbouring the KRAS G12C mutation. A total of 129 patients (59 with lung cancer; 42 with colorectal cancer; and 28 with other tumour types) were included in dose escalation and expansion cohorts. Interesting results in terms of clinical efficacy were confirmed in phase II, and those specific for pancreatic cancer patients are presented in Table 2 (38 patients)[62]. The second inhibitor adagrasib (MRTX849) has also been tested in phase I and II trials[56,63]. Among the 18 evaluable patients with PDAC, 33% showed a confirmed objective response, with a disease control rate of 81% and a median progression free survival of 5.4 months. To enhance the activity of KRAS G12C inhibitors, vertical inhibition strategies (double or triple with anti-SHP2, anti-MEK, and/or anti-EGFR) could be an interesting strategy[64,65]. Reports of promising results of combining anti-KRAS G12C and anti-EGFR inhibitors have been made concerning colorectal cancer[66,67] though not as yet with regards pancreatic cancer. However, considering that KRAS G12C mutated tumours represent only a small proportion of all PDAC patients, the development of other strategies is critical in this patient group.

Table 2 Safety and toxicity data regarding KRASG12C inhibitors in pancreatic ductal adenocarcinoma.

	Adagrasib (n = 21)[57]	Sotorasib (n = 38)[63]
Objective response rate (95%CI)	33 (15-57)	21 (10-37)
Disease control rate	81	84
Median duration of response (months)	NA	5.7 (1.6-non evaluable)
Progression-free survival (months) (95%CI)	5.4 (3.9-8.2)	4.0 (2.8-5.6)
Overall survival (months) (95%CI)	8.0 (5.2-11.8)	6.9 (5.0-9.1)
Grade 3-4 toxicities	27	16
Dose reduction	40 (solid tumours)	13
Discontinuation for toxicity	0	0

Alternative strategies have already been developed in this regard. The first involved the development of other selective inhibitors against more common codons, such as G12D. A selective and potent KRAS G12D inhibitor, MRTX1133[68], with non-covalent and picomolar binding affinity to the protein was developed by the Mirati^® team. Pre-clinical efficacy testing gave good results in both cell lines and murine models[68,69]. The second strategy involved the development of direct pan-specific KRAS inhibitors and proteolysis targeting chimeras (PROTAC), capable of sparing NRAS and HRAS[15]. The PROTACs are an emerging class of drugs that specifically degrade proteins through the cellular protein degradation system[70]. They interact simultaneously with a protein of interest and an E3 ligase, forming a ternary complex that allows the E3 ligase to ubiquitinate and induce degradation of the target protein[71]. It remains to be established whether targeting all three RAS isoforms simultaneously would be compatible with regards to therapeutic window in patients (i.e., with manageable toxicities). A third indirect inhibition strategy has been developed to interfere with nucleotide exchange and KRAS activation through the inhibition of SHP2 or SOS1[72-76].

A SPECIFIC SUBGROUP OF PATIENTS WITH KRAS WILD-TYPE PDAC TUMOURS

Patients with KRAS wild-type (KRAS_WT) tumours represent a specific subgroup of patients, accounting for approximately 10% to 15% of all PDAC cases[34,60,77,78]. In a recently published study[78], these patients showed a better prognosis than those for whom KRAS was mutated, with an OS of 24 months vs 17.5 months, despite a higher OS for the KRAS mutants in this study compared to that generally reported in this population. The KRAS_WT subtype is also characterized by a higher prevalence of actionable alterations, ranging from 30% to 40% in various studies[34,60,77]. An analysis of molecular alterations based on tumour tissue from around 2500 patients demonstrated that KRAS_WT pancreatic cancer more commonly harboured mutations in DNA damage repair genes (i.e., ATM, BAP1, BRCA2, FANCE, PALB2, and RAD50), chromatin remodeling genes (ARID1A, ARID2, KMT2C, KMT2D, PBRM1, SETD2, and SMARCA4), cell cycle control genes (CDKN2A, CCND1, and CCNE1), and BRAF[77]. Other important molecular phenotypes such as high microsatellite instability (4.7% vs 0.7%; P < 0.05) and high tumour mutation burden (4.5% vs 1.0%, P <0.05) were also more frequently observed in KRAS_WT pancreatic cancer in this study[77]. KRAS_WT PDAC harbour a distinct molecular landscape with the frequent detection of therapeutically targetable oncogenic fusions, including ALK, BRAF, FGFR2, MET, NRG1, and RET fusions, which have been reported in 19% to 67% of KRAS_WT pancreatic cancer compared with 0% to 1% of KRAS-mutated PDAC[34,77,79,80]. Patients with KRAS_WT pancreatic cancer with therapeutically targetable fusion proteins who received targeted therapy, including entrectinib or larotrectinib for NTRK fusion, afatinib for NRG1 fusion, crizotinib for MET fusion, selpercatinib for RET fusion and erdafitinib for FGFR2 fusion, have experienced durable responses[34,79-81].

One proposed strategy accounting for the difficulty in rendering routine NGS for all PDAC patients, is to first determine KRAS status through targeted sequencing on the basis of which NGS would only be performed for KRAS_WT patients[77]. The impact of identifying KRAS_WT tumours appears to exceed the greater proportion of actionable molecular alterations. Indeed, nimotuzumab, a monoclonal antibody directed against EGFR, was effective when combined with gemcitabine for wild-type KRAS tumours with an acceptable tolerance profile in Asian patients[82]. Specific treatments should therefore be developed and/or validated for these tumour subtypes.

TRANSCRIPTOMIC SIGNATURES FOR PRECISION MEDICINE IN PDAC

One team determined transcriptomic signatures, based on gene expression profiles in patient-derived organoids, primary cell cultures, and xenografts. They first correlated the transcriptomic profiles with concurrent gemcitabine sensitivity analyses, which then permitted the identification of tumours showing particular sensitivity to different treatments, including gemcitabine, oxaliplatin, 5-FU, irinotecan, and docetaxel[83-85]. A statistical approach was used to derive a multi-gene signature predictive of gemcitabine sensitivity from these preclinical models (GemPred signature)[83]. This signature was finally tested in a single-centre cohort and validated in a large multicentre cohort of resectable PDAC patients[83]. The same approach was used for metastatic PDAC and allowed good patient stratification[84]. The same team then developed a sensitivity signature for FOLFIRINOX (specifically for 5-FU, irinotecan, and oxaliplatin)[85]. A clinical trial is currently ongoing (PACsign, NCT05475366) to evaluate the strategy of guiding first-line treatment of patients based on these predictive signatures.

CONCLUSION

To conclude, the spectrum of actionable molecular alterations is evolving rapidly, with the development of KRAS inhibitors that could potentially provide actionability for almost all PDAC tumours. The higher frequency of actionable alterations harboured by KRAS_WT tumours provides numerous therapeutic opportunities. Considering the potential clinical benefit of molecularly-matched treatments, along with their improved tolerance rates compared to chemotherapies, generalizing NGS to widen therapeutic options for patients with PDAC would be expected to improve prognosis and should soon be endorsed by academic societies.