Familial Pancreatic Cancer

INTRODUCTION

Pancreatic ductal adenocarcinoma (PDAC) is the deadliest of all major cancer types with most patients having advanced disease at diagnosis. Recent data from the National Cancer Institute’s (NCI) Surveillance, Epidemiology, and End Results (SEER) registry reveal a 5-year survival rate of 3% for patients with metastatic pancreatic cancer, increasing to 14.4% for those with regional (node-positive) disease and 41.6% for those with localized disease confined to the pancreas.¹ Pancreatic cancer survival rates have increased significantly in the past decade, and the current 5-year survival is 11%. Experts predict that by 2030 PDAC will be the second-leading cause of cancer death in the United States.²

Progress is needed, not only in the treatment of this disease but also in the diagnosis and pancreatic risk assessment that would enable more individuals to pursue early detection. Despite the high mortality associated with pancreatic cancer, screening the general population for PDAC is not recommended.³ The average lifetime risk of developing PDAC is low (1.6%, 1 in 64), and more importantly, the age-specific incidence in the general population is low (approximately 1/1000 to 1/10,000, depending on age). As a result, even the best screening tests would generate hundreds to thousands of false positives for every true positive test.⁴ However, individuals with an estimated 5% or a higher lifetime risk of developing PDAC are now encouraged to undergo surveillance, as per the International Cancer of the Pancreas Screening (CAPS) Consortium recommendations.⁵ The so-called high-risk individuals (HRIs) carry either a deleterious germline mutation in one of the pancreatic cancer susceptibility genes and/or have a strong family history of PDAC, with at least one first-degree and one second-degree relative affected with PDAC.⁵

In this review, we elaborate on the inherited predisposition to familial pancreatic cancer (FPC) and focus on the role of surveillance for individuals at high risk of pancreatic cancer.

INHERITED PREDISPOSITION

Pancreatic cancer has a significant inherited predisposition.^6–8 It is well-known that having a family history of PDAC increases the risk of developing the disease. The entity of FPC refers to kindred with at least a pair of first-degree relatives with pancreatic cancer, a definition that helps quantify risk. Both the number of family members affected, and their degree relation to one another can be used to estimate this risk. For example, Klein and colleagues⁶ found that having two first-degree relatives with pancreatic cancer equates to an elevated PDAC risk ratio of 6.4-fold over average risk, corresponding to a lifetime risk of 8% to 12%. The magnitude of the risk is also associated with the age-at-PDAC of the relatives, with higher risk observed among relatives diagnosed at younger than 50 years.⁹ Further studies are needed to better define these risk estimates, including how inherited susceptibility gene variants modify the risk in families. The definition of FPC was created before the availability of widespread gene testing. Nowadays, individuals at increased risk are typically grouped according to whether they have a familial-only risk or if they have a known pancreatic cancer susceptibility gene variant.^3,10

Despite the value of obtaining a detailed cancer family history, relying on a family history of pancreatic or other cancers to identify potential gene mutation carriers will miss most mutation carriers, and most affected individuals do not come from kindred with a recognizable inherited cancer syndrome.⁸ When multigene panel testing is applied to unselected patients with PDAC, close to 10% are found to have a high-penetrance pathogenic variant in one of the known pancreatic cancer susceptibility genes,^11–13 including BRCA2 (breast cancer 2), ATM (ataxia telangiectasia mutated), BRCA1 (breast cancer 1), Lynch syndrome-associated genes, CDKN2A (cyclin dependent kinase inhibitor 2A) (responsible for Familial Atypical Mole and Multiple Melanoma)⁷ or STK11 (serine threonine kinase 11) (responsible for Peutz–Jeghers syndrome), among others (Table 1).

Table 1.

Pancreatic cancer susceptibility genes

Gene Name	Average Lifetime Risk for PDAC	Increased Relative Risk of Developing PDAC	Associated Inherited Syndromes	Other Cancers/Clinical Manifestations	Mechanism of Action
BRCA2 BRCA1	5–10%13,17 3%13,17	~4–6x13 ~2.5x13	Hereditary breast and ovarian cancer	Breast, ovarian, prostate, gastric cancers17 Breast, ovarian, prostate, colorectal, gallbladder cancers17	Homologous repair
ATM	9.5%18	~6x13	Ataxia telangiectasia	Breast, prostate cancers	DNA repair
CDK2NA	17%19,20	~12–20x13	Familial atypical mole and multiple melanoma syndrome	Melanoma	Cell cycle regulation
MLH1 MSH2 MSH6	4%21	~7x13 ~7x13 Unknown	Lynch syndrome	Colorectal, gynecologic, urothelial, brain cancers intestinal, gastric	Mismatch repair
PALB2	2–3%22	~6x13	None	Breast cancer	Homologous repair
TP53	Unknown	~7x13	Li-Fraumeni syndrome	Most human cancers	DNA repair
PRSS1	7–40%23,24	~10x13	Hereditary pancreatitis	Recurrent acute pancreatitis	Trypsin activation
STK11	11–36%25,26	~30x13	Peutz-Jeghers syndrome	Breast, GIT, gynecologic, lung cancers	AMPK signaling
CPA1 CPB1	~0.5%	~3x ~9x27	Hereditary pancreatitis Subclinical pancreatitis	None	Acinar-cell ER stress

Abbreviations: AMPK, adenosine monophosphate-activated protein kinase; CPA1, carboxypeptidase A1; CPB1, carboxypeptidase B1; ER, endoplasmic reticulum; GIT, Gastrointestinal tract; MLH1, mutL homolog 1; MSH2, mutS homolog 2; MSH6, mutS homolog 6; PALB2, partner and localizer of BRCA2; PRSS1, serine protease 1; TP53, tumor protein p53.

The significant potential of lowering cancer mortality by identifying close relatives of PDAC cases who have deleterious susceptibility gene variants has led the National Comprehensive Cancer Network^14,15 and the American Society of Clinical Oncology¹⁶ to develop guidelines recommending germline gene testing to all patients diagnosed with PDAC regardless of their age at diagnosis or family history, and if appropriate, their first-degree relatives.

For most of these pancreatic cancer susceptibility genes, deleterious variants predispose to other cancers as well.⁸ Hereditary recurrent acute pancreatitis, mostly due to deleterious variants in PRSS1 is a rare but important mechanism of pancreatic cancer susceptibility.⁸

In most populations, deleterious susceptibility gene variants are most commonly identified in BRCA2 and ATM among unselected PDAC cases, with a prevalence of approximately 2%.^11,13 Founder mutations are an important contributor in certain populations such as the common BRCA2 founder mutation in individuals of Ashkenazi Jewish descent²⁸ and the CDK2NA founder mutation in the Dutch.¹⁹

The average lifetime risk of PDAC is highest among patients with Peutz–Jeghers syndrome (11% to 36%^25,26) and those who have a germline pathogenic variant in CDKN2A (17%).^19,20 The risk of pancreatic cancer is also high for PRSS1 mutation carriers with earlier studies estimating a lifetime risk of approximately 40%. However, a more recent study estimated a much lower risk of about 7%.^23,24

The lifetime risk for BRCA2 and ATM mutation carriers is estimated to be approximately 10%. For ATM mutation carriers, this risk has been estimated to be about 1.0% by age 50 years, increasing to 6.3% by 70 years of age, and to 9.5% by 80 years.¹⁸ Among BRCA2 mutation carriers, estimates of the relative risk¹⁷ and the odds ratio (OR)¹³ are similar (approximately 6-fold overall in most studies). In contrast, the lifetime risk among BRCA1 mutations carriers is approximately 3%, below the 5% risk threshold that risk experts have used as a guideline when considering who should be screened for pancreatic cancer in the absence of a family history.

The PDAC risk among carriers with Lynch syndrome-associated germline mutations (in MLH1, MSH2, PMS2, and MSH6) is also estimated to be near the 5% threshold for surveillance,²¹ and the PDAC risk among PALB2 germline mutation carriers is thought to be similar to BRCA2, though limited data are available (see Table 1).

Recent efforts to identify additional FPC susceptibility genes have identified several genes with variants that are thought to contribute to pancreatic cancer risk by inducing pancreatic acinar cell endoplasmic reticulum stress, a mechanism known to contribute to some forms of hereditary pancreatitis.²⁷ The case-control studies^27,29 have found evidence that rare variants in CPA1 and CPB1 (functionally classified as endoplasmic reticulum [ER] stress-inducing variants genes) are more common in patients with PDAC than in non-cancer controls (OR for CPA1: 3.65 [95% CI, 1.58–8.39] and OR for CPB1: 9.51 [95% CI, 3.46–26.15]).²⁷ A recent genome-wide association study identified a common deletion variant in CTRB2 (chymotrypsinogen B2) as ER-stress inducing (OR for PDAC 1.36).³⁰ Of note, acute pancreatitis is usually not a feature of PDAC cases associated with ER stress variants.

The aforementioned risk values are subject to variation depending on the germline variant itself as well as additional inherited and environmental factors.⁸ For example, smoking is a major nongenetic factor that independently contributes to the increased risk of PDAC with OR, 3.7 (95% CI, 1.8–7.6).^6,7 Smoking is also associated with a substantially lower average age-at-diagnosis in familial PDAC kindred. In individuals with hereditary pancreatitis, smoking lowers the age of onset by approximately 20 years.³¹ Overall, pathogenic variants in the known germline pancreatic susceptibility genes known to date explain only a minor portion (less than 20%) of the familial clustering of pancreatic cancer.⁸

Polygenic risk scores (the risk of developing a disease based on the total number of changes related to the disease) are being developed for several cancer types in an effort to refine risk estimates.³² Although polygenic risk scores can refine risk and have promise for common diseases, their clinical value for less common diseases such as pancreatic cancer is currently limited, even when combined with other risk factors (family history, smoking, obesity, and so forth).³³ One study has used polygenic risk scores to estimate PDAC risk among patients with new-onset diabetes.³⁴

PANCREATIC CANCER SURVEILLANCE

For over two decades, the potential of early detection to reduce pancreatic cancer mortality has formed the basis of pancreas surveillance studies. In recent years, the accumulated evidence of these studies finds that pancreatic surveillance of HRIs can downstage tumors with significantly improved outcomes among those diagnosed with pancreatic cancer. Candidates for pancreatic surveillance should have a detailed discussion about the potential benefits and risks of surveillance and the need for long-term compliance. A patient-tailored approach should consider an individual’s family history, germline mutation status, most recent imaging findings, and genetic and environmental factors.

Goals of Surveillance

In 2020, the International CAPS Consortium issued its modified recommendations with regard to pancreatic surveillance, updating the CAPS consensus statements from 2013.³⁵ Consensus regarding the main goal of pancreatic surveillance did not change: reducing pancreatic cancer-related deaths by the detection and treatment of Stage I PDAC and precursor lesions with high-grade dysplasia.^5,36 The clearest evidence to date that pancreatic surveillance can improve long-term outcomes is reported in a study of 354 HRIs enrolled in the CAPS program, where long-term survival was common among patients who maintained surveillance compared with those whose surveillance lapsed.³⁷

Whom to Screen and When

The selection of individuals eligible for pancreatic surveillance currently relies on three key criteria: family history, germline mutation status, and age (Table 2). These criteria allow clinicians to best assess pancreatic cancer risk.⁵

Table 2.

Summary of the pancreatic cancer surveillance recommendations in high-risk individual by organization

	International CAPS Consortium5 (2020)	AGA36 (2020)	ACG38 (2015
	• All patients with Peutz-Jeghers syndrome
	• All patients with hereditary pancreatitis
	• All carriers of a germline CDKN2A mutation
Who?	• Carriers of a germline BRCA2, PALB2, ATM, MLH1, MSH2, or MSH6 gene mutation with at least one affected FDR or SDR
	• Individuals from familial pancreatic cancer kindred, defined as at least two affected FDRs who are an FDR to at least one with pancreatic cancer.
	• Carriers of a germline BRCA1 gene mutation with at least one affected FDRa		• Carriers of a germline BRCA1 gene mutation with at least one affected FDR or SDR
	• Age to begin surveillance depends on patienťs gene mutation status and family history
	• There is no consensus on the age to end surveillance
	• Start surveillance at age 40 y for all individuals with Peutz-Jeghers syndrome	• Start surveillance at age 35 y for all individuals with Peutz-Jeghers syndrome
When?	• Start surveillance at age 40 y for all carriers of a germline CDKN2A mutation and individuals with hereditary pancreatitis		• Start surveillance at age 50 y, or 10 y younger than the earliest age of PC in the family for all carriers of a germline CDKN2A mutation
	• Start surveillance at age 45 or 50 y or 10 y younger than youngest affected blood relative for all carriers of a germline BRCA2, BRCA1, PALB2, ATM, MLH1, MSH2, or MSH6 mutation	• Start surveillance at age 50 y, or 10 y younger than the earliest age of PC in the family for all carriers of a germline BRCA2, BRCA1, PALB2, ATM, MLH1, MSH2, or MSH6 mutation and for individuals with FPC kindred (without a known germline mutation)
	• Start surveillance at age 50 or 55 ya or 10 y younger than youngest affected blood relative for FPC kindred (without a known germline mutation)
How?	• MRI/MRCP and EUS at baseline
	• Alternate MRI/MRCP and EUS (no consensus if and how to alternate) during follow-up
	• Serum CA19–9 and CT only for concerning findings		• Not considered
	• Surveillance annually if no or only non-concerning abnormalities
	• Surveillance every 3 or 6 mo if concerning abnormalities for which immediate surgery is not indicated
	• Surgery if positive FNA and/or high suspicion of malignancy on imaging

FDR, first-degree relative; SDR, second-degree relative; FPC, Familial Pancreatic Cancer; MRI, magnetic resonance imaging; MRCP, magnetic resonance cholangiopancreatography; EUS, endoscopic ultrasound; FNA, fine-needle aspiration; CT, computed tomography; PC, pancreatic cancer.

^aConsensus was not reached.

Current recommendations of the International CAPS Consortium regarding family history (when there is no known susceptibility gene mutation) state that individuals are eligible for surveillance if they have at least two affected blood relatives with PDAC on the same side of the family, of whom at least one is an affected first-degree relative, who in turn has an affected first-degree relative.⁵ The recommendations regarding age-to-commence surveillance are based on the age-at-diagnosis of pancreatic cancer in HRIs. Most pancreatic cancers in individuals with a family history of PDAC are diagnosed after age 55, and on average have a later age-at-diagnosis than those with a known pancreatic cancer susceptibility variant.^11,37 The consensus at the last International CAPS Consortium was to initiate surveillance at age 50 years or later for this group, although 22.1% of the experts considered 55 years as the adequate age to begin. This was consistent with the latest American Gastroenterological Association (AGA) clinical practice update recommendations³⁶ and the American College of Gastroenterology (ACG) guidelines.³⁸

The recommendations for surveillance of those with susceptibility gene variants are specific for each gene.^5,36,38 Given their high lifetime risk of pancreatic cancer, carriers of mutations in STK11 (Peutz–Jeghers syndrome) and CDKN2A (Familial Atypical Multiple Mole Melanoma, only mutations that inactivate p16) are advised to participate in pancreatic surveillance, regardless of their family history.⁵ Initiating surveillance for CDKN2A mutation carriers is recommended at age 40 years and at age 30 to 40 years for patients with Peutz–Jeghers syndrome,⁵ with the ACG suggesting 35 years as a suitable age.³⁸ For patients with hereditary pancreatitis, owing to disease-causing variants in PRSS1, CPA1, and CTRC,⁵ surveillance is recommended after age 40 years, or 20 years after the first pancreatitis attack, irrespective of gene status.^5,31,36,38

Discriminating changes of pancreatitis from the presence of pancreatic neoplasia using current imaging tests is challenging, and for this reason, pancreas surveillance of those with a history of recurrent acute pancreatitis should be undertaken at centers of excellence.³⁶ For ATM, BRCA2, PALB2, and Lynch syndrome mutation carriers, most guidelines recommend surveillance if individuals have at least one affected first-degree blood relative with pancreatic cancer.^5,36 Consensus as to the pancreatic cancer family history criteria for BRCA1 mutation carriers (who have a low average lifetime risk of developing PDAC) was not reached among CAPS experts, although most still encourage surveillance in this group.⁵

Among mutation carriers, a family history of pancreatic cancer modifies risk, but the extent to which it does is not well-defined. In the early years of pancreas surveillance, before much evidence and experience had been accumulated, it was considered important to limit eligibility to the highest risk individuals. In recent years, groups have begun investigating the yield of surveillance among mutation carriers without a family history of PDAC. Katona and colleagues³⁹ examined this question among BRCA1, BRCA2, ATM, and PALB2 mutation carriers. The sample size was too small to adequately address this question, but two individuals (3%) developed PDAC, a similar detection rate to that reported in other studies among high-risk groups with a family history of PDAC (yielding an estimate that 111 to 135 patients will need to be screened to identify one with a PDAC or a high-grade precursor lesion).^40,41

Some studies have reported a low diagnostic yield among individuals under surveillance with familial-only risk. For example, the Dutch FPC Surveillance Study Group reported on 366 HRIs (including 201 mutation-negative FPC kindred and 165 PDAC susceptibility gene mutation carriers) and revealed a cumulative PDAC incidence of 9.3% in the mutation carriers group, as opposed to 0% in the FPC kindred one (P < 0.001).⁴² These findings led Overbeek and colleagues⁴² to question the value of surveillance in those with meeting only familial risk criteria. However, this conclusion is likely premature as this has not been the experience in the CAPS program, although not all familial risk patients undergo gene testing. Indeed, most of the risk estimates in FPC kindred were based on genetically untested families.^6,43 However, the low number of PDAC cases and the relatively young age of their cohort (whose risk of PDAC will increase with time) were limitations of their study. Abe and colleagues⁴⁴ demonstrated through a study involving 464 HRIs that the risk of neoplastic progression is higher among patients with a known deleterious germline mutation than in those with a familial risk alone [hazard ratio (HR), 2.85; 95% CI, 1.0 to 8.18; P < 0.05]. Abe and colleagues⁴⁴ showed the risk of developing PDAC among individuals with familial risk alone is still significant and recommended surveillance in this group. The latest AGA Clinical Practice Update recommended as their third best practice advice that genetic testing should be offered to FPC kindred who meet the criteria for surveillance.³⁶

Pancreas Surveillance Tests

Endoscopic ultrasound (EUS) and magnetic resonance imaging (MRI)/magnetic retrograde cholangiopancreatography (MRCP) are currently the most widely used modalities for pancreatic surveillance. Pancreatic protocol computed tomography (CT) has been shown to be useful in characterizing solid lesions after they were found on surveillance imaging,⁷ but CT is not generally used as a first-line test because of the radiation dose and intravenous contrast requirement. Furthermore, EUS and MRI/MRCP have the advantage of being superior at detecting very small pancreatic cysts.^5,45 Developments in CT, such as the use of artificial intelligence methods to improve the detection of subtle lesions, could change that recommendation in the future.⁴⁶

Several studies have attempted to compare the diagnostic yield of these imaging technologies. For example, Canto and colleagues ⁴⁷compared EUS, MRI/MRCP, and CT and found that the detection of subcentimeter pancreatic cysts was superior with EUS and MRI/MRCP. Some studies have compared overall diagnostic yield that includes cysts and solid lesions. A meta-analysis of 2,122 HRIs who underwent imaging found no significant difference between EUS and MRI in their ability to detect high-grade dysplasia, T1N0M0 PDAC, or cysts.⁴⁰ This study and prior studies^40,48,49 found that EUS can detect subtle focal parenchymal abnormalities which might result from pancreatic intraepithelial neoplasia (PanIN), although their clinical significance in the setting of surveillance has yet to be defined.

The detection of PanIN remains a major challenge of pancreatic surveillance. Most PDACs are thought to originate from PanIN, whether the individual has a family history, inherited susceptibility gene mutation, or a sporadic form of cancer. As subcentimeter pancreatic cysts with imaging characteristics of intraductal papillary mucinous neoplasms (IPMNs) are commonly observed in patients undergoing surveillance, it would be easy to conclude (incorrectly) that PDACs that emerge in those with a familial/inherited risk often go through an IPMN-like pathway. However, PDACs that do arise in the background of pancreatic cysts, often do so away from preexisting, non-worrisome pancreatic cysts, and have independent genetics.⁵⁰ Indeed, even among patients with large sporadic IPMNs, it is common to find that the pancreatic cancer is genetically distinct from the IPMN and presumed to have arisen from PanIN.⁵¹ Concomitant high-grade PanINs originating in areas of the pancreas distinct from pancreatic cysts are often found in subjects with resected sporadic low-grade IPMNs. Among HRIs, the presence of pancreatic cysts is associated with a modest increased relative risk of developing PDAC compared with those without such cysts.³⁷

Developments in imaging may one day allow for the detection of PanIN within the intact pancreas, but this remains a challenge. Some investigators are looking at pancreatic juice collected from the duodenum at the time of EUS (pancreatic juice profiling) using digital next-generation sequencing. Such profiling can reveal the molecular alterations within the pancreas, but they are not sufficient to localize PanIN.⁵² Other approaches using molecular imaging such as those targeting the integrin α_vβ6-binding peptide which has been used to image pancreatic cancer⁵³ could have utility for detecting large PanIN.

The most important question regarding the diagnostic yield of imaging tests in the setting of pancreatic surveillance is the accuracy for detecting small (1 cm or less) pancreatic cancers. Many factors contribute to diagnostic performance including clinician/radiologist expertise, imaging technique used, and the variable imaging characteristics of small tumors. The low incidence of PDAC even in high-risk cohorts has made it a difficult question to answer definitively. A multicenter blinded prospective study by Harinck and colleagues,⁴⁵ comparing the efficacy of MRI and EUS in HRIs revealed that EUS was particularly sensitive for the detection of solid lesions less than 20 mm, with two PDACs (including a Stage I PDAC) detected by EUS but not by MRI.

Outcomes of Surveillance

Few studies of pancreatic surveillance among HRIs have reported sufficient numbers of patients diagnosed with PDAC to evaluate long-term outcomes. In their long-term (16-year) CAPS study of 354 HRIs undergoing surveillance, Canto and colleagues³⁷ demonstrated that PDAC and/or high-grade dysplasia developed in 7% of the cohort (24 patients), with 71% of the PDACs detected at stages I and II. Outside of surveillance, patients with symptomatic PDAC have less than a 20% chance of having low-stage resectable disease.¹ The downstaging observed in the CAPS study was associated with better overall 3-year survival; 85% for the 10 PDACs detected in asymptomatic HRIs during surveillance versus 25% for the four symptomatic HRIs who developed PDAC after dropping out of surveillance (P < 0.0001).³⁷ Similarly, a prospective study following 178 CKDN2A/p16 Leiden mutation carriers in three expert European centers demonstrated that 75% of the PDACs detected during surveillance were resectable,¹⁹ and a much higher proportion than the 15% reported for historical controls.⁵⁴ Furthermore, the 5-year survival rate was markedly higher (24%) than that previously described for individuals with symptomatic sporadic PDAC (4% to 7%) in the Dutch Cancer Registry.⁵⁵

Early detection could be improved if it was possible to identify features on imaging or other characteristics to better tailor surveillance interval recommendations. Based on what has been estimated about the growth rate of pancreatic tumor masses⁵⁶ and the experience gained from surveillance of HRIs, most patients are recommended to undergo annual surveillance. Progression to PDAC occurs most often in those older than 60 years, those with numerous pancreatic cysts,³⁷ and in mutation carriers,⁴⁴ but these factors alone are not sufficiently discriminating of future PDAC risk to warrant changing surveillance intervals.

When worrisome features are detected on imaging, surveillance intervals should generally be adjusted.³⁶ The experience of surveillance of sporadic pancreatic cysts that led to the Sendai and Fukuoka International Consensus Guidelines for management of mucinous cysts are helpful: cyst size ≥ 3 cm, thickened/enhancing cyst walls, mural nodule in the cyst or main pancreatic duct (MPD), MPD dilation greater than 5 mm, abrupt change in MPD caliber, suspicious cytology for pancreatic malignancy, or rapid cyst growth rate greater than 2 mm in 6 months or greater than 4 mm in 1 year.⁵⁷ The ACG clinical guidelines acknowledge the difficulty in establishing clear-cut inclusion criteria for surgery and state that the decision should be individualized after multidisciplinary evaluation by experts of the field.³⁸ In fact, Dbouk and colleagues⁵⁸ assessed the diagnostic performance of the Fukuoka and CAPS guidelines for the management of pancreatic cysts in HRIs and showed that both would have failed to adequately recommend surgery for PDAC patients. They found that the Fukuoka criteria had a lower sensitivity (40%) for selection of HRIs for surgery compared with individuals with sporadic cysts, missing 60% of cysts with invasive carcinoma or IPMNs with high-grade dysplasia. The CAPS criteria also missed 40% of resected neoplastic IPMNs. Both consensus criteria might have recommended unnecessary surgery for 15% of HRIs.⁵⁸ One reason for the poor ability of imaging characteristics to predict the emergence of PDAC, especially in the familial/inherited risk setting, is the inability of current imaging tools to detect PanIN with high-grade dysplasia.

In the early years of pancreatic surveillance, patients were often sent for pancreatic resection for worrisome features, or other concerning abnormalities. For example, in their meta-analysis including 1,551 familial HRIs, Paiella and colleagues⁵⁹ reported a pooled proportion of overall surgery of 6% (95% CI 4.1–7.9, P < 0.001) and unnecessary surgery of 68.1% (95% CI 59.5–76.7, P < 0.001), with final pathology reports revealing diagnoses incompatible with screening goals in most cases. In addition, a review of such cases found that up to 15% of solid pancreatic lesions were benigns.⁵⁵ Nowadays, clinicians have learned to be more selective.

In summary, pancreatic surveillance of HRIs has been shown to detect PDAC with a high-resectability rate. Surgical resection in this setting is associated with acceptable morbidity, minimal mortality, and remarkable long-term survival.⁶⁰ There is some evidence that participation of HRIs in a screening program did not lead to an increase in cancer worry, as could have been previously presumed.⁶¹ Instead, taking part in a surveillance plan procured a sense of control and even reduced anxiety in some patients.⁶² The evaluation of long-term outcomes of HRIs should continue, but not without acknowledging that compliance to a surveillance program remains a territory that requires substantial exploration. Longitudinal studies should aim at identifying factors associated with poor surveillance compliance in HRIs.⁶³ In the United States, cost can be a barrier to being able to undertake annual EUS or MRI/MRCP surveillance, as insurance coverage for pancreatic surveillance is variable. Indeed, across the United States, among individuals diagnosed with PDAC (NCI SEER data-general population cases), having insurance is associated with a lower stage at diagnosis.⁶⁴

THE POTENTIAL OF BIOMARKER TESTS TO IMPROVE PANCREAS SURVEILLANCE

Remarkable efforts have been made to identify HRIs who qualify for pancreas surveillance in the hope of optimizing early detection of pancreatic cancer in this unique population. Novel strategies are needed to better stratify PDAC risk in HRIs with the aim of increasing the number of eligible patients for pancreatic cancer surveillance.⁶⁵ Likewise, new tests are being evaluated to improve detection of high-grade precancerous lesions and early-stage invasive carcinomas not only in HRIs but also in the general population.

Improving the diagnostic accuracy of circulating tumor markers such as CA19–9 would be of great value. As a marker, CA19–9 is significantly influenced by common variants in the genes responsible for its synthesis (FUT2 and FUT3). Abe and colleagues⁶⁶ found that classifying individuals into CA19–9 reference ranges based on these FUT2/FUT3 variants significantly improved the accuracy of CA19–9 (and other tumor markers) for diagnosing pancreatic cancer. They showed that a tumor marker gene test combined with a CA19–9 test enhanced its diagnostic performance (particularly in patients with intact FUT3), with 66.4% sensitivity and 99.2% specificity for subjects with localized PDAC. To have the greatest impact, these tests need to detect stage I PDAC. A major challenge for the detection of stage I PDACs with a blood-based test is that smaller cancers generally shed fewer biomarkers into the blood-stream. Studying biomarkers from patients with stage I PDACs is similarly challenging because these cases are very rare,⁶⁶ and many newly diagnosed patients now receive neoadjuvant chemotherapy, obscuring assessment of their pathologic stage at presentation.⁶⁷

Liquid biopsies have been used to evaluate circulating tumor DNA (ctDNA) as a noninvasive tool to detect early-stage pancreatic cancer. For pancreatic cancer detection, initial studies relied on mutated KRAS ctDNA. More recent studies have used pan-cancer ctDNA tests as part of multi-cancer detection approaches.⁶⁸ Further evidence is needed before considering its implementation in a clinical setting.

SUMMARY

Studies in recent years have provided a deeper understanding of the inherited predisposition to PDAC and its relationship to FPC. Experts have identified at present at least a dozen genes known to contribute to the development of PDAC in mutation carriers. Patients whose PDAC is localized to the pancreas have a better chance of survival than those with a more advanced form of the disease.^64,69 Individuals with familial and inherited risk of pancreatic cancer who undertake pancreas surveillance and are diagnosed with early-stage PDAC, especially stage I disease, have an excellent chance to achieve long-term survival.

KEY POINTS.

• First-degree relatives of individuals with familial pancreatic cancer and individuals with a deleterious germline mutation in one of the pancreatic cancer susceptibility genes are at significantly increased risk of developing pancreatic cancer and can potentially benefit from surveillance.

• Surveillance focuses on the early detection of pancreatic ductal adenocarcinoma and its high-grade precursor lesions using endoscopic ultrasound examination and/or magnetic resonance imaging/magnetic resonance cholangiopancreatography.

• Patients undergoing pancreas surveillance are more likely to be diagnosed with low-stage resectable pancreatic cancers and to achieve long-term survival.

• Novel strategies are needed to better stratify pancreatic cancer risk. Emerging tests are currently being evaluated, which could eventually improve the detection of high-grade precancerous lesions and early-stage invasive pancreatic cancer.

CLINICS CARE POINTS.

• Pancreatic ductal adenocarcinoma (PDAC) has an overall poor prognosis. Symptoms associated with PDAC commonly manifest late in the course of disease.

• Eligibility criteria for pancreas surveillance are based on age, germline mutation status, and family history. High-risk individuals (HRIs) with an estimated 5% or higher lifetime risk of developing PDAC are advised to undergo surveillance for PDAC.

• The accumulated evidence of pancreas surveillance studies demonstrates that pancreatic surveillance of HRIs can result in the diagnosis of early-stage PDAC, and this is associated with improved outcomes.

• Endoscopic ultrasound and/or magnetic resonance imaging/magnetic resonance cholangiopancreatography are considered the standard imaging modalities for screening to date. Computed tomography scan has been shown to be useful in characterizing concerning solid lesions found on surveillance but is not generally used as a first-line test.

• Blood-based biomarkers and circulating tumor DNA tests represent promising noninvasive tools to detect early-stage pancreatic cancer.