#90241: Pancreatic Cancer

Health professionals' implicit biases shape behaviors, communications, and interactions, which then produce differences in diagnoses and ultimately treatments and interventions. Implicit biases are subtle and unconscious and may unwittingly produce professional behaviors, attitudes, and interactions that reduce patients' trust and comfort with their provider.

Racial and socioeconomic differences in surgical intervention rates, treatment at high-volume hospitals/centers, and morbidity and mortality rates have been noted, with the largest disparities between Black (and to a slightly lesser extent Hispanic) and White Americans [23]. Several factors are implicated, but implicit biases and insurance status are identified as potentially modifiable contributors.

Cigarette smokers have at least a two-fold greater risk for pancreatic cancer than nonsmokers. The risk increases directly with daily number of cigarettes consumed and duration of smoking. In heavy smokers with polymorphism in the carcinogen-metabolizing enzyme gene glutathione S-transferase theta 1 (GSTT1), the risk is up to five-fold greater [25,26].

Excess risk decreases with smoking cessation. The risk of pancreatic cancer among current smokers (relative risk: 2.5) decreased 48% two years after smoking cessation, and within 10 to 15 years after cessation, it approximated that of nonsmokers [26].

In the United States, estimates indicate that 11% to 32% of deaths from PDAC are attributable to tobacco smoking. It is estimated that cessation of smoking could eliminate up to 25% of pancreatic cancer deaths [24,26].

Limited evidence suggests alcohol consumption may be associated with risk of developing PDAC, but findings of population-based studies are inconsistent. In pooled cohort data of 1.5 million light, heavy, or never-drinkers, heavy drinkers had a greater relative risk of developing PDAC than never-drinkers (relative risk: 1.29) or light drinkers (relative risk: 1.36). Light drinkers had no difference compared to never-drinkers (relative risk: 0.96) [27].

Most studies have assumed additivity between average effects of smoking and alcohol and oversimplified their impact on burden of pancreatic cancer. However, the combined effect of smoking and total alcohol intake on risk of PDAC is likely non-additive. It appears that only heavy consumption of liquor (but not wine or beer) increases the risk of PDAC in ever-smokers [27].

Obesity (defined as a mass index [BMI] >30) during early adulthood is associated with a greater risk of certain cancers and earlier onset. Several studies have reported that the incidence of early-onset pancreatic cancer has been increasing in parallel with the expanding prevalence of obesity. Other obesity-related cancers increasing in incidence and early-onset include colorectal cancer, multiple myeloma, and cancers of the uterus, gallbladder, and kidney. An analysis of incidence data from 25 state cancer registries covering 14.7 million invasive cancers diagnosed between 1995 and 2014 found the incidence of PDAC was increasing among adults 25 to 49 years of age, with steeper rises in successively younger birth cohorts [135]. Diets high in processed meat, high-fructose beverages, and saturated fat were associated with obesity, diabetes, and PDAC. Obesity is associated with a 20% to 40% higher mortality rate from PDAC, and obesity at an older age is associated with lower overall survival [13; 28].

Although BMI is widely used as a marker for general adiposity, measures of visceral obesity show a stronger correlation with metabolic syndrome, insulin resistance, and certain gastrointestinal (GI) malignancies. The proximity to visceral organs and drainage via the portal system may explain the strong correlation of inflamed visceral adipose tissue (VAT) in obese subjects with metabolic dysfunction and pancreatic cancer [29].

There is some evidence that higher consumption of red/processed meat is associated with elevation in pancreatic cancer risk, but other studies have failed to identify dietary risk factors for PDAC [11]. Pancreatic cancer incidence may be lower in persons with higher intake of fresh fruits and vegetables rich in folate and lycopenes (e.g., tomatoes) [30].

A link between vitamin D and risk for pancreatic cancer is inconsistent, but some data suggest low plasma 25-hydroxyvitamin D levels may increase the risk for pancreatic cancer, especially in those with low retinol/vitamin A intake [31]. Coffee and tea consumption are not associated with pancreatic cancer risk, despite early reports to the contrary [24].

Numerous studies and meta-analyses have found systemic/nonmodifiable factors that increased the relative risk, hazard ratio, or odds ratio of developing pancreatic cancer. These include individuals with greater height (relative risk: 1.81); individuals with blood groups A, AB, and B (hazard ratio: 1.32, 1.51, and 1.72, respectively); and patients with hepatitis B infection (odds ratio: 1.50) or systemic lupus erythematosus (hazard ratio: 1.43). Biologic explanations for some of these associations are not yet understood, and some data may have potential confounders. Infectious etiologies warrant more investigation [11,32].

Periodontitis describes a chronic inflammatory response to a disease-associated, multispecies bacterial community in the subgingival region. Periodontal disease is associated with pancreatic cancer, even when controlling for gender, smoking, BMI, diabetes, and alcohol consumption [33]. The inflammatory processes of periodontitis occur locally, but systemic dissemination of inflammatory mediators, subgingival species, and bacterial components contribute to digestive cancers (including PDAC) by activating proinflammatory pathways, inducing gene expression related to cell proliferation, apoptosis, and immune responses linked to carcinogenesis, cell migration, invasion, and metastasis [34].

A high-risk subgroup for PDAC are patients with chronic pancreatitis, often secondary to chronic alcohol use disorder, smoking, hypertriglyceridemia, diabetes, or renal failure [2]. Patients with chronic pancreatitis show a 26-fold increase in risk of developing PDAC. This risk increases with duration. Among patients with chronic pancreatitis of 20 years' duration, approximately 5% will progress to PDAC.

Concomitant smoking enhances the risk of neoplastic progression [2,35]. Hereditary pancreatitis further increases the risk of pancreatic cancer by more than 50-fold. In these individuals, the cumulative risk of pancreatic cancer by age 70 years is 40% [24].

Pancreatic cancer has complex relationships with diabetes and obesity that are only recently becoming understood. A population cohort study underscored the complex relationship between metabolic abnormalities and PDAC. Glycemic status, insulin resistance, and hyperinsulinemia were independently associated with an increased risk of pancreatic cancer mortality, even in individuals without diabetes [36].

The association between pancreatic cancer and diabetes was noted as early as 1833, clearly documented by the 1930s, and characterized in a large cohort of patients with pancreatic cancer from Mayo Clinic in 1958 [37]. Several meta-analyses have greatly refined the risk-factor status of diabetes.

Long-standing (i.e., more than five years) diabetes (both type 1 and type 2) is associated with increased risk of developing PDAC [13]. The overall risk for PDAC increases 4- to 7-fold in those with diabetes of a duration less than three years [38]. The relative risk associated with diabetes levels off after five years, with a 1.5-fold greater risk [39]. Increased baseline hemoglobin A1C (HbA1C) levels correlate with subsequent development of PDAC [40].

Long-standing diabetes modestly increases the risk of PDAC, which decreases with diabetes duration [11,37]. The initial three-year period after diabetes diagnosis is high risk for PDAC, as confirmed by prospective pancreatographic screening [41].

With diabetes medications, insulin use has been associated with increased risk of PDAC, but this finding is attributed to reverse causality [11,42]. Metformin use in patients with diabetes and PDAC was associated with improved two-year survival (30.1% vs. 15.4%) and median overall survival (15.2 months vs. 11.1 months) in patients without metastases [43]. One metformin study reported negative findings [44].

Long-standing diabetes in patients who develop PDAC is associated with significantly lower overall survival (14.4 months vs. 21.7 months) and significantly higher mortality (harm ratio: 1.52) compared with patients without diabetes who develop PDAC [11,45].

Diabetes of the exocrine pancreas (formerly type 3c diabetes) is the second most common type of new-onset diabetes in adults (behind type 2 diabetes) [42]. Acute or chronic pancreatitis is one of the most prevalent risk factors for PDAC and the most frequent cause of diabetes of the exocrine pancreas. Pancreatitis leads to postpancreatitis diabetes mellitus in up to 83% of patients [42]. In a registry study involving 139,843 individuals, the proportion of pancreatic cancer was 3.1% among those with postpancreatitis diabetes mellitus, compared with 2.3% in those with type 2 diabetes followed by pancreatitis, 2.0% in those with pancreatitis alone, and 0.6% in individuals with type 2 diabetes alone [42].

Numerous clinical series have identified new-onset diabetes, weight loss, and soft tissue changes in patients with PDAC at diagnosis, but their inter-relationship and connection to PDAC remained unaddressed. From 2000 through 2015, temporal changes in the five years preceding PDAC diagnosis of 219 patients diagnosed with PDAC were compared to 657 controls [46]. From 60 to 30 months before PDAC diagnosis, patients did not significantly differ from controls. However, starting at 30 months prediagnosis, PDAC showed three distinct metabolic phases, each marked by onset and significant progressive worsening of one or more metabolic abnormalities [46]:

Based on evidence of increases in body temperature before PDAC diagnosis, browning and loss of subcutaneous abdominal tissue is estimated to begin 18 months before PDAC. Browning of white abdominal tissue is a mechanism of subcutaneous abdominal tissue loss in cancer; its purpose is to generate heat [46].

Symptoms of cachexia and muscle loss (e.g., anorexia, fatigue, reduced exercise tolerance) appear shortly (less than six months) before PDAC diagnosis. The onset of objective weight loss precedes PDAC diagnosis by one year or more. New-onset diabetes appears a median of six to nine months before PDAC diagnosis [46].

Cancer cachexia is a paraneoplastic syndrome characterized by pronounced weight loss and muscle wasting triggered by cancer-induced systemic inflammation [47]. Cachexia develops in about 80% of patients with PDAC during the disease course, often before the tumor is clinically apparent. Cachexia negatively impacts treatment response and survival, and one-third of patients with PDAC die from cachexia-associated complications, including impaired immunity and cardiopulmonary dysfunction. No curative treatments exist [47].

Pancreatic cancer-associated diabetes mellitus might be a major contributor to PDAC-induced cachexia. The co-occurrence is frequent, and the relationship between pancreatic cancer-associated diabetes and PDAC-induced cachexia was clarified in a 2020 study [47]. Compared with patients without pancreatic cancer-associated diabetes, those with pancreatic cancer-associated diabetes did not have a higher risk of cachexia, a greater degree of weight loss, or lower skeletal muscle mass. Among patients with cachexia, weight loss and skeletal muscle mass were comparable between patients with and without pancreatic cancer-associated diabetes. Fasting blood glucose levels and PDAC-derived diabetogenic factors did not correlate with weight loss or muscle mass or predict cachexia in patients with pancreatic cancer-associated diabetes. A notable finding was the consistently high prevalence of cachexia and muscle wasting regardless of tumor size and stage in PDAC [47]. These results argue against pancreatic cancer-associated diabetes and hyperglycemia in mediating PDAC-induced cachexia.

Cancer cachexia is characterized by systemic inflammation with resultant skeletal muscle breakdown and increased circulating amino acids to support tumor growth. Pancreatic cancer-associated diabetes is a metabolic strategy by PDAC to fuel tumor growth. PDAC cells have a high demand for glucose (termed "glucose addiction"); hyperglycemia promotes invasion and migration of PDAC cells. PDAC-induced cachexia and pancreatic cancer-associated diabetes are distinct metabolic reprogramming induced by PDAC cells to secure amino acids and glucose for tumor growth [47].

Unexplained weight loss/cachexia is a clue to occult PDAC, but a modality that can identify PDAC-induced cachexia is needed to take advantage of this screening opportunity [47]. Optimizing glycemic control may not alleviate weight loss or muscle wasting, and therapies targeting mediators of pancreatic cancer-associated diabetes may not protect against the development of cachexia [47]. Management of cachexia in patients with PDAC is discussed in detail later in this course.

PDAC develops in PanINs through a specific process [56]. First, mutational KRAS activation initiates pancreatic carcinogenesis. With tumor suppressor inactivation, cancer progresses. CDKN2A or SMAD4 are implicated in locally destructive disease; TP53 is involved in metastatic seeding; and concurrent SMAD4 and TP53 are often present in locally or metastatic dominant disease. IPMNs and MCNs often share the driver gene mutations and sequence of PanINs, but also show specific patterns.

Lesions involving the main pancreatic duct have a higher malignant potential than those in the branches, with the risk of malignancy at around 62% [11]. More than 90% of IPMNs are marked by activating mutations in the oncogene GNAS and/or inactivating mutations in the tumor suppressor gene RNF43[48,53,54]. GNAS mutation causes constitutive activation of adenylyl cyclase, with downstream effects driving proliferation. RNF43 encodes E3 ubiquitin-protein ligase, which functions as a tumor suppressor in the Wnt-signaling pathway. After the initiating oncogene mutation, the progression of IPMN resembles PanIN.

The risk of malignant transformation in MCNs is <15%, most often associated with an RNF43 gene mutation event (50%). As in PanINs, genetic changes accumulate with higher grade of dysplasia and invasiveness [48,53,54].

Rather than being uniformly aggressive, PDAC exhibits clinical (e.g., variable patient survival) and disease (e.g., variable chemotherapy sensitivity) heterogeneity [64,65]. The first whole-genome description of PDAC in 2008 prompted great effort to advance a patient-tailored precision medicine approach that could better address this heterogeneity. Genetic alterations and molecular subtypes in PDAC were characterized and published. PDAC was shown mutationally dominated by the four driver genes and homogeneous. In general, the findings importantly informed the biology and familial predisposition of PDAC.

However, by 2019 it was apparent that PDAC disease heterogeneity cannot be explained by genetic mutations alone, and non-genetic mechanisms, including epigenetics and the tumorigenic microenvironment, were the path forward [21,56,59,62,64,65,66,67].

Epigenetic Factors

Broadly speaking, epigenetic changes influence gene expression, without altering the DNA sequence, through modifications of DNA or chromatin structures [4]. In PDAC, these include DNA methylation and non-coding RNAs (ncRNAs).

Gene expression in PDAC can be silenced through non-mutational inactivation by aberrant promoter methylation, including the driver gene p16/CDKN2A[49]. Aberrant ncRNA expression plays a considerable role in initiation, proliferation, and chemo-resistance of PDAC. Oncogenic microRNA-21 promotes both cell proliferation and apoptosis and targets negative regulators of KRAS, which further enhances signaling by this oncogene [50,54].

Pancreatic Tumor Microenvironment

PDAC is comprised of cancer cells within dense fibrotic stroma consisting of extracellular matrix and non-neoplastic (e.g., fibroblastic, vascular, immune) cells [3]. Also described as PDAC fibrosis, the stroma makes up most of the tumor mass. This network of neoplastic tissue and stroma (the pancreatic tumor microenvironment) acts as physical barrier to drug penetration and has elements that impede host immune surveillance and antagonize host anticancer immune responsiveness, facilitating PDAC growth, survival, and treatment failure [21,51,68]. Current anti-cancer therapy includes novel approaches designed to enhance sensitivity of PDAC to the host immune system and reverse the immunosuppressive effects of the tumor microenvironment [134].

Pancreatic cancer progresses in tandem with extensive stromal deposition of extracellular matrix, recruitment and activation of cancer-associated fibroblasts, and high interstitial fluid pressures, which compresses blood vessels, causing localized hypoperfusion, reduced vascularity, and tissue hypoxia [21,69]. Extracellular matrix remodeling biomechanically induces intracellular signaling and tumor-stellate cell crosstalk. PDAC cells signal to stellate cells and recruit macrophages and immune suppressor cells. In turn, stellate cells secrete factors that promote PDAC cell proliferation, migration, and suppression of apoptosis [51]. Biochemical activation of signaling pathways regulates PDAC cell survival and promotes tumor growth and metastasis. Down-stream effects include immunosuppression, disease progression, epithelial-mesenchymal transition (a key step of the metastatic cascade) and invasive potential, and chemotherapy resistance [3,21,69].

Exosomes (a macromolecule involved in RNA degradation) released by PDAC cells accumulate in other tissues to create a premetastatic niche by activating stellate cells and inducing remodeling of the host extracellular matrix, which facilitates cancer cell invasion and growth [59,69].

Several genetic syndromes are associated with specific genetic alterations with an increased risk for pancreatic cancer (Table 3) [48,54]. Germline mutations in familial atypical multiple mole melanoma syndrome (CDKN2A) and Li-Fraumeni syndrome (TP53) are core gene drivers in sporadic PDAC. Peutz-Jeghers syndrome is caused by germline inactivation of STK11, a tumor suppressor gene. Somatic STK11 mutations are observed in approximately 4% of pancreatic cancers, suggesting STK11 inactivation plays a role in both sporadic and familial forms [49].

An estimated 10% to 15% of all pancreatic cancers are attributable to genetic causes. Pancreatic cancer aggregates in some families; 5% to 10% of individuals with pancreatic cancer have a family history of the disease [26,70,72]. Familial pancreatic cancer represents 90% of all hereditary PDAC cases. The relative risk of PDAC increases with the number of affected first-degree relatives.

A specific gene defect responsible for familial pancreatic cancer has not been identified, but a rare autosomal-dominant gene may be responsible, putting 0.4% to 0.7% of the population at risk for developing PDAC [26,70,72]. Details about the relative and lifetime risks of PDAC, and the other prevalent cancers associated with specific germline mutations in cancer susceptibility syndromes and familial pancreatic cancer, are summarized in Table 4.

Abdominal pain is the most common symptom, usually insidious in onset and often present for one to two months at the time of presentation, the pain is often severe, and unrelenting in nature. The typical gnawing, visceral quality of pain is generally epigastric, radiating to the sides and/or straight through to the back; some patients may describe the pain as originating in the back. Nighttime pain is often the predominant complaint. Some patients note increased pain after eating and worsened pain when lying flat [24,81]. Rarely, acute pain develops when an episode of acute pancreatitis results in tumor occlusion of the main pancreatic duct [84].

While roughly one-third of patients may not have pain at the time of initial presentation, all patients will develop pain at some point [24]. Pancreatic cancer is one of the most painful malignancies, and effective pain control is extremely important [85]. This issue will be discussed in detail later in this course.

The most characteristic sign of tumor in the pancreatic head is obstructive jaundice, for which patients may seek medical attention before their tumor grows large enough to cause abdominal pain (and thus, a somewhat better prognosis). These patients usually notice a darkening of their urine and/or lightening of their stools before they or their families notice the change in skin pigmentation. Jaundice secondary to a tumor in the body or tail of the pancreas typically occurs at a later stage and may be secondary to liver metastases of PDAC [2,84].

Pruritus can accompany and often precedes obstructive jaundice. If present, it is often the patient's most distressing symptom [24].

A characteristic feature of pancreatic cancer, significant weight loss may be related to cancer-associated anorexia and/or subclinical malabsorption from pancreatic exocrine insufficiency caused by pancreatic duct obstruction. Nausea and early satiety from gastric outlet obstruction and delayed gastric emptying from the tumor can contribute to weight loss [24]. Significant weight loss is a symptom of cachexia.

Pancreatic cancer cachexia is a multifactorial paraneoplastic syndrome characterized by a loss of skeletal muscle mass, commonly associated with adipose tissue wasting and anorexia, fatigue, and reduced exercise tolerance. Cachexia develops in approximately 80% of patients with PDAC, in whom the syndrome is typically present at diagnosis and responds poorly to therapeutic interventions [47,86].

Pancreatic cancer leads to the development of cachexia through a combination of distinct factors that explain its high prevalence and clinical importance in this disease [86]:

The initial assessment can uncover additional diagnostic clues. Undiagnosed diabetes leads to symptoms of glucose intolerance (e.g., polyuria, polydipsia). PDAC can interfere with production of digestive enzymes by the pancreas (pancreatic exocrine insufficiency) and with the ability to break down food and absorb nutrients (malabsorption) in some patients. This malabsorption causes bloating, gas, and a watery, greasy, and/or foul-smelling diarrhea, leading to weight loss and vitamin deficiencies [81].

While long-standing diabetes is a risk factor for later development of PDAC, new-onset hyperglycemia or diabetes has been identified in most patients at diagnosis of otherwise asymptomatic PDAC. Deregulation in glucose homeostasis is often accompanied by changes in subcutaneous adipose tissue. Both represent paraneoplastic syndromes caused by the underlying PDAC [2].

This research is among the most important knowledge advances in PDAC in the past decade. In addition to metabolic deregulation, the pre-diagnostic soft tissue changes and symptoms of cachexia have profound implications for screening, early diagnosis, treatment selection, and patient prognosis [2].

Tumors can also grow locally into the duodenum (proximal for the head of the pancreas, distal for the body and tail of the pancreas) and result in an upper gastroduodenal obstruction [13]. Tumor in the body or tail of the pancreas may cause splenic vein obstruction, resulting in splenomegaly, gastric and esophageal varices, and gastrointestinal hemorrhage [81].

The American Society of Clinical Oncology recommends a multiphase computed tomography (CT) scan of the abdomen and pelvis using a pancreatic protocol or magnetic resonance imaging (MRI) be performed for all patients with pancreatic cancer to assess the anatomic relationships of the primary tumor and to assess for the presence of intra-abdominal metastases.

(https://ascopubs.org/doi/10.1200/JCO.19.00946 Last Accessed: August 27, 2024)

Strength of Recommendation/Level of Evidence: Strong/high

Multidetector computed tomography (MDCT) angiography with intravenous (IV) contrast is the preferred imaging for initial evaluation of suspected PDAC. The Pancreatic CT Protocol standardizes its use, making MDCT highly accurate for assessing tumor extent, vascular invasion, and distant metastases [11,16,88,89]. The degree of contact between the tumor and local blood vessels (i.e., uninvolved, abutted, or encased) is used to define the most optimal initial treatment [134]. The NCCN recommends that MDCT angiography should also cover the chest and pelvis for complete staging [11].

MDCT is 77% accurate in predicting resectability and 93% accurate in predicting unresectability [85]. MDCT may be superior to magnetic resonance imaging/magnetic resonance cholangiopancreatography (MRI/MRCP) in vascular enhancement of a PDAC, the most important parameter of resectability. However, MDCT is inferior to MRI/MRCP in depicting isodense tumors or tumors smaller than 1.5 cm in size [54].

Abdominal MRI/MRCP with IV contrast also employs a standard multiphase protocol in PDAC, with efficacy comparable to MDCT in preoperative evaluation and assessment of vascular invasion. The sensitivity of MRI/MRCP in detecting liver metastases is nearly 100% (vs. 80% with MDCT) [81,85].

Selection of initial MDCT or MRI/MRCP is typically based on local availability and expertise [81,85]. Following initial MDCT, MRI/MRCP is used when PDAC is highly suspected but negative on MDCT, for characterizing small or indeterminate pancreatic and hepatic tumors, and in patients with severe allergy to iodinated IV contrast material used in MDCT [54,81,85].

With endoscopic retrograde cholangiopancreatography (ERCP), contrast dye is injected into the biliary ducts and pancreatic duct with an endoscope, and the level of obstruction is delineated. In some case, placement of a biliary stent can help relieve symptoms of jaundice [85]. Patients with obstructive jaundice may have ERCP as the first diagnostic procedure [81].

Transabdominal ultrasonography is useful in initial screening of patients who present with possible obstructive jaundice and can rapidly and accurately assess for biliary obstruction. However, definitive diagnosis requires other imaging [24].

Endoscopic ultrasonography is superior to MDCT in detecting solid pancreatic lesions less than 2 cm in size, with accuracy of about 92% [54]. Endoscopic ultrasonography-guided fine-needle core biopsy (preferred over fine-needle aspiration) is recommended to obtain a histologic diagnosis and to provide material for molecular testing [134].

With the restricted field of view, endoscopic ultrasonography is complimentary to MDCT, but it should be used before other imaging options if no pancreatic mass is evident on MDCT. Endoscopic ultrasonography is also valuable in detecting tumor involvement of blood vessels or lymph nodes [11,89].

Positron-emission tomography (PET) imaging alone does not offer added advantages to MDCT. Combining PET with CT (PET/CT) is a more recent development that may enhance the detection of occult metastases in pancreatic cancer. The NCCN guidelines consider PET/CT an evolving technology; its role in the diagnosis of PDAC is not yet established [11].

Around 5% to 10% of the population lacks the enzyme necessary to produce CA19-9; monitoring pancreatic cancer with this marker will not be possible in these individuals [24]. Biliary obstruction also stimulates the secretion of CA19-9. Hyperbilirubinemia is associated with elevated CA19-9 and false positivity in patients with obstructive jaundice. Following the treatment of obstruction, re-evaluation of CA19-9 should improve its diagnostic utility [92].

The NCCN recommends measurement of serum CA19-9 levels after neoadjuvant treatment, prior to and immediately following surgery before adjuvant therapy, and in surveillance. The importance is stressed of obtaining CA19-9 immediately before a therapeutic intervention to have an accurate baseline from which to follow response [11].

The AJCC system (Table 6) is used for staging PDAC in two contexts [16,94]:

Clinical staging identifies the primary tumor and its vessel involvement, enlarged or suspicious lymph nodes, and metastatic disease sites. TNM staging provides important prognostic information (Table 7) but does not assess whether the PDAC tumor is amenable to surgical resection [54,94].

Complete resection is the only potentially curative treatment for PDAC, but fewer than 20% of patients presenting with PDAC have localized and easily resectable tumors, and noncurative resections provide no survival benefit. Thus, accurate assessment of resectability is crucial [24,87,89].

The NCCN guideline classes PDAC resectability into the following clinical stages [11]:

Localized PDAC falls on a spectrum from high to low resectability, determined by the extent of vessel contact and whether the involvement is arterial or venous (Figure 1) [11,54,84,87,89,95]. Major peripancreatic vessels include the superior mesenteric vein and artery, portal vein, common hepatic artery, and celiac artery. Tumor contact can be characterized as encasement (≥180 degrees of the vessel circumference), abutment (<180 degrees of the circumference), or direct involvement (absence of fat plane between tumor and vessel).

SPECTRUM OF PANCREATIC CANCER RESECTABILITY

Figure 1

Source [84,89,95]

In the past, vascular infiltration by PDAC was considered unresectable, but surgical advances have increased the number of patients with initial borderline resectable or locally advanced disease who can undergo resection. In general, venous abutment or encasement is usually borderline resectable as long as the venous segment is reconstructable. Arterial reconstruction is substantially more difficult and riskier than venous reconstruction with comparable tumor contact.

Based on PDAC clinical status of resectable, borderline resectable, locally advanced, or metastatic disease, additional considerations and therapeutic approaches will be undertaken. The time-urgency between the first availability of full imaging findings, multidisciplinary evaluation, the diagnostic and staging workup, discussion with the patient of available treatment options, and treatment initiation cannot be overstated in this aggressive malignancy.

For patients with resectable or borderline resectable PDAC, neoadjuvant therapy consists of chemotherapy with or without radiation therapy before radical pancreatic resection [99]. Radical pancreatic resection may include Whipple procedure (pancreaticoduodenal resection) or total pancreatectomy when necessary for adequate margins. Distal pancreatectomy is indicated for tumors of the body and tail of the pancreas. Following surgical resection, patients may receive postoperative chemotherapy or postoperative chemoradiation therapy [99].

Chemotherapy with or without targeted therapy is recommended for patients with locally advanced PDAC [99]. For patients without metastatic disease, this should be followed by chemoradiation therapy. If removal is a possibility, radical pancreatic resection may be attempted. Palliative surgery options include surgical biliary and/or gastric bypass, percutaneous radiologic biliary stent placement, or endoscopic biliary stent placement.

Treatment of metastatic or recurrent PDAC is limited to chemotherapy with or without targeted therapy [99]. Palliative approaches should be used whenever available and feasible to improve patient comfort and quality of life.

Pancreaticoduodenectomy (Whipple Procedure)

Used for tumors in the pancreatic head or periampullary region, the conventional Whipple procedure involves removal of the pancreatic head, duodenum, gallbladder, and the antrum of the stomach, with surgical drainage of the distal pancreatic duct and biliary system, usually through anastomosis to the jejunum. The primary reason for removing so much of the intra-abdominal structures is that they all share a common blood supply [24,102].

The former high morbidity and mortality rates of Whipple have declined with the greater experience of a more limited number of surgeons who regularly perform the procedure in high-volume centers [102]. Common morbidities include delayed gastric emptying in roughly 25% of patients, which may require nasogastric decompression and a longer hospital stay. Pancreatic anastomotic leak can be treated with adequate drainage. Postoperative abscesses are not uncommon [24].

With operative mortality associated with Whipple decreasing from around 25% in the 1970s to less than 2% at high-volume centers in the 2010s, the focus has shifted from surviving the operation to surviving the cancer [104].

Distal Pancreatectomy

Distal pancreatectomy is a procedure for tumors in the pancreatic body or tail. It has a lower mortality than standard Whipple, but its use in curative resection is limited; with tumors in this location seldom causing bile duct obstruction, most patients present at a later stage with unresectable disease. The procedure involves resection of the distal pancreas containing the tumor with splenectomy and over-sewing of the distal pancreatic duct. Complications involve pancreatic stump leak, hemorrhage, and endocrine insufficiency. Laparoscopic exploration should precede attempted resection, because occult peritoneal metastases are common [16,24].

Total Pancreatectomy

Total pancreatectomy, the least commonly performed procedure with the highest associated mortality (8.3%), may be needed to achieve an R0 resection margin for tumors in the neck of the pancreas, especially with extension into the body or tail, and in multifocal PDAC. Total pancreatectomy may be an option to pancreatic anastomosis in highly selected patients with a high-risk pancreas (small pancreatic duct) and obese patients with pancreatic fat infiltration. The metabolic consequences of permanent exocrine insufficiency and diabetes have a detrimental impact on quality of life and long-term survival [16,24,102].

Vascular involvement has traditionally been a formal contraindication to resection. With recent advances, venous resection and reconstruction can achieve R0 resection with similar overall survival and morbidity compared to surgery without venous resection. However, arterial resection during Whipple is associated with increased mortality and morbidity (bowel ischemia, hemorrhage, thrombosis) and is generally not recommended [16].

Progress in neoadjuvant therapies may downstage some tumors with arterial invasion to borderline resectable or resectable disease, making resection more achievable. Despite these advancements, it is currently accepted that arterial reconstruction is only appropriate in highly selected patients in high-volume centers with surgeons who are familiar with the advanced techniques required for reconstruction [16].

Total pancreatectomy should be considered in patients with locally advanced tumors who undergo pancreatectomy with arterial resection and reconstruction [16].

In most patients with jaundice, early resection without biliary drainage is preferred. Preoperative drainage is indicated in patients with cholangitis or with obstructive jaundice scheduled for neoadjuvant therapy. Endoscopic retrograde placement of a fully covered metal stent is preferred. Endoscopic ultrasonography-guided stent placement is an effective and safe alternative [16].

According to the American Society of Clinical Oncology, all patients with resected pancreatic adenocarcinoma who did not receive preoperative therapy should be offered six months of adjuvant chemotherapy in the absence of medical or surgical contraindications. The mFOLFIRINOX regimen is preferred in the absence of concerns for toxicity or tolerance.

(https://ascopubs.org/doi/10.1200/JCO.19.00946 Last Accessed: August 27, 2024)

Strength of Recommendation/Level of Evidence: Strong/high

In addition to single chemotherapy agents, the FDA has approved regimens of these agents, including FOLFIRINOX (consisting of folinic acid [also referred to as leucovorin], fluorouracil [5-FU], irinotecan [IRN], and oxaliplatin [OX]) (Table 9) [3,24,80,99]. Available chemotherapies are associated with acute and delayed toxicities, some of which can be dose-limiting (Table 10). Table 11 summarizes the 2021 NCCN guideline for chemotherapy and chemoradiotherapy in PDAC.

Fluoropyrimidines

Fluorouracil is a fluorinated (fluoro)-pyrimidine antimetabolite that inhibits thymidylate synthase and interferes with RNA synthesis and function, with some effect on DNA.

Capecitabine is an oral fluoropyrimidine that undergoes hepatic hydrolysis to form fluorouracil. The final enzyme, thymidine phosphorylase, is present at higher levels in tumor tissue, providing better selectivity and tolerability.

Gemcitabine is a pyrimidine antimetabolite that inhibits DNA polymerase and ribonucleotide reductase, which in turn inhibit DNA synthesis, blocks DNA replication and several forms of DNA repair [3,24,80,99].

Erlotinib

Erlotinib is a human epidermal growth factor receptor type 1/epidermal growth factor receptor (HER1/EGFR) tyrosine kinase inhibitor. EGFR is expressed on the cell surface of normal cells and cancer cells. Erlotinib inhibits intracellular phosphorylation, which prevents further downstream signaling, resulting in cell death [3,24,80,99].

Paclitaxel

Paclitaxel protein bound is a microtubular inhibitor (albumin-conjugated formulation) and a natural taxane that prevents depolymerization of cellular microtubules, which results in DNA, RNA, and protein synthesis inhibition [3,24,80,99].

Irinotecan Liposomal

Irinotecan and its active metabolite SN-38 bind reversibly to the topoisomerase-1 DNA complex and prevent re-ligation of the single-strand breaks, leading to exposure time-dependent double-strand DNA damage and cell death. Irinotecan liposomal is used in combination with fluorouracil and leucovorin [3,24,80,99].

Platinum agents (e.g., cisplatin, oxaliplatin) and olaparib are recommended in patients with mutation in DNA damage repair (DDR) genes by the NCCN. DDR mutations are present in up to 24% of PDACs, most commonly BRCA1/2 and PALB2. Germline BRCA1/2 mutations (gBRCAm) affect approximately 7% of patients with PDAC [108]. DDR genes encode for proteins in the homologous repair pathway and DNA double-stranded break repair; thus, mutations may be more sensitive to further DNA damage [99].

Cisplatin inhibits DNA synthesis by the formation of DNA cross-links; denatures the double helix; covalently binds to DNA bases; and disrupts DNA function. Oxaliplatin is an alkylating agent. Following intracellular hydrolysis, the compound binds to DNA, forming cross-links that inhibit DNA replication and transcription, resulting in cell death [24,99].

PDACs with DDR mutations demonstrate improved responses to platinum-based therapies, and patients with advanced PDAC showed significantly improved median overall survival (22 months vs. 9 months) compared with nonplatinum therapy [96].

Poly (ADP-ribose) polymerase (PARP) inhibition has been posited to act synergistically with BRCA1/2 mutations by inhibiting single-stranded break repair, causing an accumulation of DNA damage and tumor-cell death [99,109]. Olaparib is a PARP inhibitor FDA-approved for PDAC with gBRCAm as maintenance therapy to sustain a progression-free state during platinum-based chemotherapy in metastatic PDAC [96].

The NCCN expands the use of olaparib to PDAC with gPALB2m. There are calls to expand these agents to PDACs with somatic DDR mutations [108].

The approved indications for the following agents are biomarker-defined, rather than by tumor site (e.g., pancreatic).

Pembrolizumab

Pembrolizumab is indicated in patients with microsatellite-instability-high (MSI-H) or dMMR mutations. Immune checkpoint inhibitors (ICIs) have efficacy in solid tumors with a high tumor mutational burden, and MSI-H or dMMR mutation solid tumors are associated with high tumor mutational burden. The ICI pembrolizumab is an anti-programmed death receptor-1 antibody that releases inhibition of the immune response, improving antitumor immunity [11,96].

Pembrolizumab is approved for any solid tumor with MSI-H or dMMR mutation that progresses during treatment without any satisfactory alternative treatment options [11,96]. This agent represented the first FDA approval (in 2017) with a biomarker-defined indication (i.e., agnostic of cancer site) [107]. Although this mutation is present in only about 1% of PDAC tumors, up to 83% of patients with dMMR PDAC respond to pembrolizumab [110].

Larotrectinib and Entrectinib

Larotrectinib and Entrectinib are neurotrophin receptor kinase (NTRK) inhibitors approved (in 2018 and 2019) for advanced, morbid, or unresectable solid tumors with NTRK fusion mutations, found in less than 1% of PDCAs [96].

The mutation product, TRK fusion protein, activates mitogen activated protein kinase-extracellular regulated kinase and phosphoinositide3 kinase-serine threonine signaling pathways, implicated in the oncogenesis of pancreatic cancer [96]. The NCCN recommends larotrectinib and entrectinib as first-line and subsequent treatment options for patients with NTRK gene fusion-positive locally advanced or metastatic PDAC [11].

A variety of data on chemotherapy efficacy are available, allowing for comparison of available agents in specific patient populations (Table 12). However, the terminology used can be confusing. Disease-free survival and progression-free survival are synonymous terms, and choice of the term used in this section will reflect the reference material. This is also the case with median survival and median overall survival. Unless noted otherwise, all patient outcomes are reported as median figures.

The CONKO-001 trial established gemcitabine as standard adjuvant chemotherapy. In this study, 354 patients were randomized to receive gemcitabine or observation after resection and followed a median 136 months. Gemcitabine led to a 24% improvement in overall survival, a 10.3% absolute improvement in 5-year survival (20.7% vs. 10.4%), and a 4.5% improvement in 10-year survival (12.2% vs. 7.7%), compared to observation [111,112].

The ESPAC-3 trial showed the importance of completing the full post-resection adjuvant chemotherapy course (six cycles) in extending median overall survival of these patients compared with those not completing chemotherapy (28.0 months vs. 14.6 months) [96].

A continuation, ESPAC-4, found adding another fluoropyrimidine-based agent (capecitabine) to gemcitabine was superior to gemcitabine alone in median survival (28.0 months vs. 25.5 months) and five-year survival (28.8% vs 16.3%). A synergistic effect between gemcitabine and capecitabine on the DNA thymidylate enzyme was suggested [96].

PRODIGE-24 randomized 493 patients (ECOG performance status ≤1) with resected PDAC to modified FOLFIRINOX or gemcitabine for 24 weeks. At median 33.6 month follow-up, the disease-free survival with modified FOLFIRINOX was 21.6 months, compared with 12.8 months with gemcitabine [113]. Grade 3/4 toxicities were more frequent with mFOLFIRINOX (75.9%) than gemcitabine (52.9%). Nonetheless, the median 54.4-month overall survival with resection followed by mFOLFIRINOX is the longest survival reported to date with phase III results [5,114]. Thus, modified FOLFIRINOX is recommended as standard adjuvant chemotherapy in patients with excellent functional status; either gemcitabine/capecitabine or gemcitabine alone should be considered for individuals with poorer functional status [134].

Tolerance of adjuvant therapy remains a limitation, and patients commonly receive less than 50% of the planned dose, reflecting exposure to significant chemotherapy-related toxicity in patients experiencing substantial post-resection morbidity [2].

First-Line Chemotherapy in Metastatic PDAC

5-FU has been used in pancreatic cancer treatment since the 1950s. Patients with advanced PDAC typically show response rates greater than 20% and median survival of 2.5 to 6 months [24,80].

In 1997, gemcitabine replaced 5-FU as first-line treatment in metastatic PDAC by improving one-year survival rates (18% vs. 2%) and median overall survival (5.65 months vs. 4.41 months) [32]. Subsequently, numerous attempts to improve gemcitabine efficacy in metastatic PDAC have involved adding another cytotoxic drug [2,96]. Some show marginal but statistically significant improvements in median survival over gemcitabine alone (Table 13).

The NCIC CTG PA.3 trial found a nonmeaningful clinical improvement with gemcitabine/erlotinib over gemcitabine alone in median overall survival (6.24 months vs. 5.91 months). Despite FDA approval for locally advanced/metastatic PDAC, the clinical impact of this modest gain with increased toxicity can be questioned [32,96].

PRODIGE 4/ACCORD 11 demonstrated that patients with advanced PDAC and ECOG performance status ≤1 had better outcomes with FOLFIRINOX than gemcitabine in median overall survival (11.1 months vs. 6.8 months) and progression-free survival (6.4 months vs. 3.3 months). Following these findings, FOLFIRINOX became standard first-line therapy for candidate patients [2].

FOLFIRINOX was associated with more toxicities, but the six-month degradation in quality of life was better in FOLFIRINOX than gemcitabine (31% vs. 66%). Improved cancer control with FOLFIRINOX may be due to the inclusion of irinotecan, which has activity against PDAC and synergistic activity when given prior to 5-FU [96].

The MPACT study demonstrated an improvement of 1.8 months in both median overall survival and median progression-free survival with gemcitabine plus nab-paclitaxel versus gemcitabine alone, leading to another first-line option for metastatic PDAC [96].

In February 2024, the FDA approved irinotecan liposome with oxaliplatin, fluorouracil, and leucovorin (NALIRIFOX), for first-line treatment of metastatic pancreatic adenocarcinoma [140]. Efficacy was demonstrated in NAPOLI 3, a randomized, multicenter trial comparing NALIRIFOX with gemcitabine plus nab-paclitaxel in 770 patients previously untreated in the metastatic stage of disease. Treatment efficacy was superior in the NALIRIFOX arm for overall survival (11.1 months vs. 9.2 months) and for progression-free survival (7.4 months vs. 5.6 months) [140]. The most common adverse reactions of NALIRIFOX therapy (occurring n at least 20% of patients) were diarrhea, fatigue, nausea, vomiting, decreased appetite, abdominal discomfort, mucosal inflammation, and pancytopenia. NCCN guidelines now include NALIRIFOX in its list of preferred regimens for treatment of patients with metastatic pancreatic cancer [141].

The NAPOLI 3 trial results suggest that NALIRIFOX may have advantages over FOLFIRINOX for treatment of metastatic PDAC. The median progression-free survival was one month longer than that reported for FOLFIRINOX in the 2011 PRODEDGE 4 trial. A systematic review and analysis of seven clinical trials (2,581 patients) testing first-line therapy for metastatic pancreatic cancer found that therapeutic efficacy (progression-free and overall survival) was similar for NALIRIFOX and FOLFIRINOX, although both were superior to gemcitabine/nab-paclitaxel [142]. NALIRIFOX was associated with higher rates of severe diarrhea but less hematologic toxicity and lower incidence of peripheral neuropathy. No clinical trial directly comparing FOLFIRINOX to NALIRIFOX have been reported. The liposomal form of irinotecan in the NALIRIFOX regimen makes it considerably more expensive than FOLFIRINOX.

Second-Line Chemotherapy in Metastatic PDAC

Second-line therapy primarily consists of doublet therapy using the alternative pyrimidine backbone to what was used in the first-line setting. In 2016, the NAPOLI-1 trial demonstrated that after progression on a first-line gemcitabine-containing regimen for metastatic PDAC, 5-FU/leucovorin plus nanoliposomal irinotecan improved overall survival from 4.2 months (with 5-FU/leucovorin alone) to 6.1 months. As with nab-paclitaxel, improving the delivery of traditional chemotherapies may lead to more effective treatments for individuals with pancreatic cancer [32].

The POLO trial examined targeted maintenance therapy in a biomarker-selected population. In patients with metastatic PDAC harboring germline BRCA1/2 mutations who had not progressed on first-line platinum-based chemotherapy, those randomized to olaparib had improved median progression-free survival (7.4 months compared with 3.8 months with placebo), but olaparib did not improve median overall survival [109]. The median duration of response to olaparib was 6 months but was more than 24 months in a subset of patients (23%), which is exceptional in metastatic PDAC [108].

In second-line chemotherapy after progression on a first-line regimen, there is considerable heterogeneity in the survival of patients, and predicting which patients will benefit is not established. The decision to pursue second-line chemotherapy should be individualized and based on the patient's goals and preferences. Factors influencing the choice of second-line therapy include the regimen used for first-line therapy, performance status and comorbidity, and mutation status [106].

The NCCN recommends neoadjuvant therapy for patients with resectable or borderline resectable tumors. Treatment at or coordinated through a high-volume center is preferred, when feasible, and participation in a clinical trial is encouraged. The preferred neoadjuvant options are FOLFIRINOX with or without subsequent chemoradiation, or gemcitabine plus albumin-bound paclitaxel with or without subsequent chemoradiation [11]. For patients with BRCA/PALB2 mutations, the preferred regimen is gemcitabine plus cisplatin (two to six cycles) with or without subsequent chemoradiation [11].

ASTRO guidelines for neoadjuvant chemoradiation specify a radiation dose of 4,500–5,040 cGy in 180–200 cGy fractions [12]. They recommend delivery of radiation therapy following two to six months of chemotherapy.

Locally advanced pancreatic cancer accounts for 30% of newly diagnosed cases. With local involvement of adjacent critical blood vessels and presence of occult micrometastatic disease, locally advanced pancreatic cancer is generally considered surgically unresectable and incurable, and the standard of care is as for metastatic disease [2].

However, the increased use of preoperative multiagent chemotherapy followed by chemoradiation has significantly expanded the pool of patients with locally advanced pancreatic cancer eligible for resection with curative intent, significantly improving the resectability and overall survival of these patients [117].

In a single-institution phase II trial, 49 patients with locally advanced pancreatic cancer received eight cycles of FOLFIRINOX followed by 50.4 Gy of photon radiation with capecitabine and losartan. Of these patients, 39 were brought to the operating room, 34 (69%) had their cancer removed, and of these, 30 patients (88%) had an R0 resection. Among patients who underwent resection, median progression-free survival and overall survival were 21.3 and 33 months, respectively, versus the 11- to 12-month historical overall survival [118].

Neoadjuvant therapy is associated with a downstaging-to-resection rate of greater than 30% in selected patients with locally advanced pancreatic cancer, with survival comparable to or better than initially resectable disease. For patients with arterial involvement, arterial divestment shows a lower morbidity and mortality rate than arterial resection and reconstruction [117].

Following neoadjuvant therapy, a restaging evaluation with pancreatic protocol MDCT is required to image tumor shrinkage and rule out local progression for resectability. However, post-neoadjuvant therapy imaging is not a reliable indicator of resectability due to its inability to distinguish post-treatment fibrosis from residual viable tumor [117]. Post-neoadjuvant therapy CA19-9 levels are predictive of tumor regression and should be used to guide decisions about suitability for surgical exploration for resection. Diagnostic laparoscopy should be routinely used to minimize nontherapeutic surgery rates [117].

After resection of pancreatic cancer following neoadjuvant FOLFIRINOX, the benefit of adjuvant chemotherapy on overall survival is unclear. Although randomized controlled trial confirmation is needed, a 2020 multicenter, retrospective study provided informative results [119]. Of 520 patients (median age: 61 years; 53.7% male) who received a median of six neoadjuvant cycles of FOLFIRINOX, 343 (66.0%) received adjuvant chemotherapy. Adjuvant chemotherapy was FOLFIRINOX for 68 patients (19.8%), gemcitabine-based chemotherapy for 201 (58.6%), capecitabine for 14 (4.1%), a combination or other agents for 45 (13.1%), and unknown for 15 patients (4.4%). The median overall survival was 38 months after diagnosis and 31 months after surgery. No survival difference was found for patients who received adjuvant chemotherapy compared with those who did not (29 months in both groups).

In multivariable analysis, the interaction of lymph node stage with adjuvant therapy was statistically significant. In patients with pathology-proven node-positive disease, adjuvant chemotherapy was associated with improved overall survival (26 months vs. 13 months). For those with node-negative disease, adjuvant chemotherapy was not associated with improved survival (38 months vs. 54 months). These results suggest that adjuvant chemotherapy after neoadjuvant therapy FOLFIRINOX and resection of pancreatic cancer was associated with improved survival only in patients with pathology-proven node-positive disease [119].

With surgical resection considered the only potentially curative option but most patients harboring unresectable PDAC tumor, nonoperative local treatment options that can provide a similar benefit are needed. Emerging radiation techniques that address organ motion have enabled curative radiation doses delivered in patients with inoperable disease [121].

In one 2021 report, patients with locally advanced pancreatic cancer were treated with hypofractionated ablative radiation therapy, using respiratory gating, soft tissue image guidance, and other methods to address organ motion and limit the dose to surrounding luminal organs [121]. At baseline, 119 patients with locally advanced pancreatic cancer and median CA19-9 level >167 U/mL received four months of induction chemotherapy, followed by ablative radiation therapy. The median overall survival from diagnosis and ablative radiation therapy were 26.8 and 18.4 months. The 12- and 24-month overall survival following therapy were 74% and 38%, and the 12- and 24-month cumulative incidence of locoregional failure were 17.6% and 32.8% [121]. Postinduction CA19-9 decline was associated with improved locoregional control and survival. Grade 3 upper GI bleeding occurred in 10 patients (8%), with no grade 4 to 5 events. This cohort study of patients with inoperable locally advanced pancreatic cancer found that ablative radiation therapy following multiagent induction therapy was associated with durable locoregional tumor control and favorable survival [121].

Symptomatic biliary obstruction develops in approximately 80% of patients with carcinoma of the head of the pancreas [143]. Biliary stents relieve troublesome pruritis and may be used to improve liver function sufficiently to permit safe administration of cancer treatment. Endoscopic retrograde stenting is superior to surgical or percutaneous approaches to address bile duct obstruction because of a more favorable adverse event rate. Successful biliary stent placement can be achieved in more than 90% of patients, with potential post-procedure complications of pancreatitis, bleeding, or cholangitis in 5% of cases [143]. Self-expandable metal stents are preferred over plastic stents in patients with a life expectancy of more than three months in terms of patency duration, less therapeutic failure and need for reintervention, lower cholangitis incidence, and better patient quality of life. Patency rates between covered and uncovered metal stents are not significantly different [16]. Endoscopic ultrasonography-guided biliary drainage is an alternative if endoscopic biliary stent placement is unsuccessful or technically not feasible.

Pancreatic cancer invasion into the duodenum can lead to secondary gastric outlet obstruction and intractable nausea and vomiting. The choice of treatment intervention depends on the functional status and patient's predicted length of survival [143]. In patients with gastric outlet obstruction, endoscopic duodenal stenting allows a quick resumption of oral intake, with a low complication rate and a short recovery period. However, the need for reintervention is higher after duodenal stenting compared with that of palliative surgery. Endoscopic ultrasonography-guided gastrojejunostomy is an effective and safe alternative to surgery [16].

Ascites in patients with metastatic PDAC may be due to peritoneal metastases. In patients with locally advanced tumors, ascites may be caused by portal vein thrombus if the tumor compresses the portal vein locally [122].

Patients with malignant ascites from pancreatic cancer can experience abdominal discomfort, nausea, vomiting, and dyspnea from the pressure of the fluid against the anterior abdominal wall and diaphragm. For most patients, survival is short, and the focus is symptom control. Symptom relief from intermittent paracentesis tends to be short-lived, and the procedure must be repeated for symptom relief. If reaccumulation requires more than once-weekly paracentesis, placement of a long-term drainage catheter is an option; complication rates are higher with indwelling catheters. Diuretics such as spironolactone and furosemide decrease the absorption of water and sodium in the kidneys and may provide some symptomatic relief [122].

The mainstay of pain management is opioid therapy, and palliation of pain can often be successfully achieved by opioid analgesics alone [122]. The core principle is a "pain/analgesia ladder," with escalation in management based on symptom severity. Initial treatment may consist of non-opioid medications (e.g., non-steroidal anti-inflammatory drugs [NSAIDs]), followed by mild/moderate opioids (e.g., tramadol, codeine), then stronger opioids (e.g., morphine, oxycodone, fentanyl) for unremitting and severe pain. At each level of pain management, adjunctive therapy (e.g., cannabinoid, ketamine, clonidine, benzodiazepine, gabapentin, pregabalin, nortriptyline, duloxetine) can be incorporated for potential additive effect [122,143].

Patients with moderate-to-severe pain should receive doses adequate to provide relief. Concern about addiction should not be a barrier to effective pain control; even with dose escalation, addiction is seldom a problem in patients with PDAC and the risk is lower than generally assumed in non-malignant pain [81,123]. Given the ongoing concerns regarding opioid misuse in the United States, drug diversion may be a consideration.

For patients with persistent nausea and vomiting for whom taking oral medications is difficult, pain control may be achieved using transdermal patches when adipose tissue is sufficient for transdermal absorption [122]. When pain is constant rather than intermittent, long-acting oral (e.g., morphine, oxycodone, oxymorphone) or transdermal (e.g., fentanyl, buprenorphine) preparations may work better [81]. Breakthrough pain can be treated with rapid-onset transmucosal or intranasal fentanyl formulations. Methadone may be advantageous in many patients and can be used in small doses as add-on to existing opioid treatment. Methadone should only by prescribed by clinicians who are familiar with the complex pharmacology and adverse effect profile of this opioid [123].

Laxatives should be considered for all patients on opioid analgesia for PDAC pain, because constipation is a nearly universal side effect. There is considerable individual variation in both efficacy and side effects. Not all patients benefit from or tolerate opioids. A trial of an alternative opioid may also be indicated. Cases of poor pain control or intolerable pain may benefit from continuous opioid infusion via epidural or intrathecal catheters [81,123].

Near the end of life, pain management for advanced and terminal PDAC can become very challenging, and an interdisciplinary approach including palliative care specialists is needed. It is important wherever possible to consider the preferences of the patient. A range of supportive care measures can be offered, including intensive home support, home care with parenteral opioids, patient-controlled analgesia, and palliative sedation [123].

Celiac plexus neurolysis offers medium-term relief, but other procedures (e.g., splanchnicectomy) are also available. Adjunctive treatments for pain, depression, and anxiety as well as radiotherapy, endoscopic therapy, and neuromodulation may be required. Palliative chemotherapy may provide pain relief as a collateral benefit [123].

Interventional therapies may be beneficial for severe pain refractory to opioids. A celiac plexus block uses a local injection of corticosteroids or a long-acting analgesic to reduce pain by disrupting visceral pain afferent pathways from the pancreas and surrounding structures. A celiac plexus neurolysis is permanent destruction of the plexus, equally effective and preferable for patients with a short life expectancy as it may provide symptomatic relief for three to six months [143].

The celiac plexus is a dense network of nerves that innervates the upper abdominal organs. Pain may be relieved by inhibiting synaptic pathways within the plexus by chemical destruction of the pathways and ganglia using dehydrated alcohol. Celiac plexus neurolysis is performed under endoscopic ultrasonography guidance [122].

Celiac plexus neurolysis improves analgesia and quality of life and decreases opioid requirements. The analgesic effect seems to vanish after eight weeks, and in most patients, pain recurs after three months. Repeated celiac plexus neurolysis benefits about 30% of patients and is normally not offered [123].

Splanchnicectomy may disrupt more nerve pathways than celiac plexus neurolysis and is a better option when there is a large mass in the region of the celiac plexus. Splanchnicectomy is seldom performed in patients with PDAC despite some evidence of long-lasting pain relief and few complications in observational series, possibly because the expertise is not widely available [123].

External beam radiation therapy with or without concomitant chemotherapy may also significantly alleviate pain due to local invasion of pancreatic cancer, frequently with improvement in cachexia and obstructive symptoms. However, it may take several weeks to achieve its maximal effect. When pain is caused by liver or bone metastases, patients may benefit from radiation therapy [16,122].

A constellation of disproportionate loss of lean body mass, weight loss, muscle wasting, adipose tissue reprogramming, and anorexia, cancer-related anorexia/cachexia syndrome (CACS) is more frequent in patients with PDAC than in any other malignancy due to the complex metabolic profile of pancreatic cancer [2]. In a study of 390 patients with advanced cancers, the rate of cachexia was highest in PDAC (89%), followed by gastric cancer (76%) and esophageal cancer (53%) [125].

Unlike simple starvation, which is characterized by a caloric deficiency that can be reversed with appropriate feeding, the weight loss of cachexia cannot be adequately treated with aggressive feeding [126]. The physical impact of CACS contributes to decreased patient quality of life, treatment response, and survival due to gross alterations in protein metabolism, increased oxidative stress, and systemic inflammation. The psychological impact also contributes to decreased quality of life for both patients and their families [125].

In CACS, an abnormally accelerated resting energy expenditure increases muscle protein breakdown and lipolysis, which seems related to activation of cytokines (e.g., tumor necrosis factor-alpha, interleukin 6 and 1 beta), and tumor-derived, potentially cachexia-inducing factors that target skeletal muscle gene products [122,126].

Potentially Beneficial Agents

Cachexia does not respond to nutritional support. There are no FDA-approved medications for treatment of CACS, and positive pharmacotherapy response in patients with anorexia associated with non-malignant disease has been difficult to translate into benefit for patients with cancer [127,128].

Many agents have been evaluated for the treatment of CACS, but only corticosteroids (e.g., dexamethasone) and progesterone analogs (e.g., megestrol acetate) have a proven benefit in the anorexia associated with this syndrome [122]. Selection is based on life expectancy and assessment of risks versus benefits. Dexamethasone is suggested for patients for whom only weeks of therapy are anticipated, while megestrol acetate or medroxyprogesterone acetate (another progesterone analog) are suggested for patients with longer life expectancies [126].

A phase III study randomized 190 patients with advanced cancer and anorexia to megestrol acetate (480 mg/day), dexamethasone (4 mg/day), or placebo for up to four weeks. Differences in primary endpoint (at least 25% improvement in appetite) between megestrol (79.3%), dexamethasone (65.5%), and placebo (58.5%) were non-significant. Hyperglycemia and deep vein thromboses were more frequent with dexamethasone than megestrol or placebo. No other differences from placebo were found [127].

In this trial, the higher rate of deep vein thromboses with dexamethasone was unexpected. Megestrol acetate is associated with thromboembolic events and is contraindicated in patients with VTE. Dexamethasone has the potential to reduce cancer-related fatigue and elevate mood, at the significant cost of accelerating catabolic effects on muscle [127]. The primary benefits associated with these drugs are increased appetite and weight gain, not improved survival, and both drugs are associated with potential harms [122].

Mirtazapine is well-known for promoting weight gain. A placebo-controlled randomized trial found that appetite scores increased similarly with mirtazapine (15 mg at night) and placebo during the 28-day study. Mirtazapine was associated with significantly less increase in depressive symptoms and higher prevalence of somnolence than placebo, but no other differences were found [128].

The evidence of benefit in patients with CACS is inconclusive for androgens and selective androgen receptor modulators, anamorelin, cyproheptadine, long-chain omega-3 fatty acids, vitamins, minerals, and other dietary supplements, nonsteroidal anti-inflammatory drugs (NSAIDs), thalidomide, and combination approaches [126]. However, a trial of low-dose olanzapine (5 mg/day) is reasonable, particularly for patients who have concurrent nausea and/or vomiting unrelated to chemotherapy or radiation therapy [126].

Cannabis and Cannabinoids

In the cannabis plant, delta9-tetrahydrocannabinol (THC) and cannabidiol (CBD) are the best-characterized therapeutic constituents. Pharmaceutical cannabinoid products containing THC (dronabinol), a THC analog (nabilone), or THC:CBD in an oromucosal spray (nabiximols, investigational) were examined for efficacy in CACS and palliative care in two meta-analyses [126].

Unfortunately, no benefit beyond placebo was found for pharmaceutical cannabinoid products in CACS, despite their superior weight gain and appetite effects in patients with advanced HIV [129]. Cancer patients with more than 30% decrease in pain with cannabinoids compared with placebo approached significance [129].

In both meta-analyses, available studies of smoked cannabis in CACS did not meet evidence thresholds and were excluded. This limits the ability to inform real-world clinical practice, where patient preference, self-titration to tolerability/effect, access, and other factors favor smoked/vaped cannabis over single-molecule pharmaceutical cannabinoids [130].

Counseling and Support

The substantial loss of body mass can cause significant distress to patients. Although advanced cachexia is irreversible, palliating anorexia in patients with advanced cancer is best approached by focusing on stimulating appetite, supporting each person's food preferences, and avoiding prescriptive dietary advice [127].

Providing education to patients and their caregivers is crucial. The objective is to promote a shared understanding about changed goals of care, and to help reduce the distress caused by reduced oral intake [127].

Counselling of family members regarding what to expect with disease progression is important—it can alleviate distress that leads to well-meaning but futile attempts to pressure or coerce the patient into increased feeding. Key points to discuss with patients and their family members, related to interactions about nutrition and eating near the end of life, include the following [131]:

Caregivers should be advised that it may be best to listen to and support the patient in a variety of other ways (such as giving the patient a massage or applying a lip moisturizer) instead of trying to talk them into eating more. Referral to a registered dietitian may provide patients and caregivers with additional opportunities to discuss concerns and challenges related to nutrition, appetite, and meal planning.

The presence of diabetes has been associated with higher mortality in patients with PDAC; corticosteroids can induce or exacerbate diabetes in these patients. For patients with PDAC-related diabetes, nutritional management by an experienced dietitian is essential [16]. Metformin or insulin is used as a first-line therapy. Insulin is often the preferred agent because of its efficacy, flexibility, and safety.

Careful monitoring of plasma glucose levels two hours after meals is widely recommended. The limited literature on this topic recommends maintaining blood glucose levels to avoid hypoglycemia and reduce symptoms of hyperglycemia.

A contributory factor to extreme weight loss may be pancreatic exocrine insufficiency, which leads to maldigestion, fat malabsorption, and steatorrhea. This complication may be caused by gradual tumor encroachment on pancreatic parenchyma or obstruction of the main pancreatic duct, or as a complication of surgery and/or irradiation. The main clinical manifestation is weight loss and malnutrition, and nonspecific symptoms such as abdominal cramping, flatulence, and urgency to defecate. Fat malabsorption does not become evident until pancreatic lipase secretion falls below 10% of normal levels [122]. The characteristic fatty stools associated with steatorrhea (loose, greasy, foul-smelling) may not be evident because patients tend to limit fat ingestion.

Pancreatic exocrine insufficiency is very frequent (>90% with tumors in the pancreatic head), and is associated with higher mortality in patients with unresectable PDAC. Pancreatic enzyme replacement therapy (PERT) improves survival in these patients [16]. Given its high incidence, diagnostic testing is not necessary. Patients suspected of fat malabsorption should be treated empirically with oral PERT [122].

The classical approach to patients with pancreatic exocrine insufficiency was restricting fat intake (<20 gm/day) to reduce steatorrhea. However, this further restricts the intake of fat-soluble vitamins, which are already malabsorbed in patients with pancreatic exocrine insufficiency and is not recommended. Frequent low-volume meals and avoidance of foods that are difficult to digest (e.g., legumes) are generally recommended [122].

Pancreatic exocrine insufficiency is treated with capsules of porcine pancreatic enzymes (pancrelipase). There are several commercial products available, and the amount of enzyme per capsule varies [81]. Doses are in United States Pharmacopeia (USP) units or International Units (IU); 90,000 USP is equivalent to 30,000 IU [122]. A healthy pancreas produces about 900,000 USP of lipase in response to a meal. Sufficient fat absorption can be maintained at around 10% of normal capacity; thus, roughly 90,000 USP per meal is needed. Because non-resected patients retain some pancreatic function, a starting dose of 75,000 USP with main meals and 25,000 with snacks should suffice in reducing steatorrhea and preventing weight loss. Enzymes are most effective when taken across the course of a meal. Following Whipple, patients will require 90,000 USP with meals and 45,000 USP with snacks [124].

Acidic gastric pH is normally neutralized by pancreatic bicarbonate secretion, which is absent in many patients with PDAC, especially following Whipple resection. Acid-suppressing therapy with a proton pump inhibitor is often required, as failure to neutralize gastric acid inactivates the enzymes [16,124].

Despite recommendation from expert groups, including the NCCN, evidence suggests PERT is underutilized. This was examined in a large commercially insured U.S. population from 2001–2013. Among patients with PDAC (32,461), 1.9% had diagnostic testing for exocrine insufficiency, 21.9% filled a prescription for PERT, and 5.5% were prescribed an adequate dose (defined as ≥120,000 USP lipase daily) [132].

Testing and appropriate dosing is infrequent and inconsistent in an insured U.S. population. Efforts are needed to educate medical providers on the best practices for managing exocrine pancreatic insufficiency in these patients [132].