Advancements in Pancreatic Cancer Screenings

November 14, 2023 - According to the National Cancer Institute, a subset of the National Institutes of Health, in 2023, an estimated 64,050 people in the United States received a new diagnosis of pancreatic cancer, comprising roughly 3.3% of all new cancer cases. Despite being a rarer cancer, it has contributed to around 50,550 deaths in 2023, accounting for 8.3% of all cancer deaths. Considering the fatal nature of the condition, researchers and clinicians are working on advancing pancreatic cancer screenings to improve early disease detection and disease survival.

Pancreatic Cancer

According to a 2023 review published in Cell, pancreatic cancer is the third leading cause of cancer-related death in the US, exceeding rates for breast cancer. By 2040, the condition is expected to surpass colorectal cancer and become the second leading cause of cancer-related deaths in the US.

A 2023 systematic review in Biomedicines notes that while pancreatic cancer is a rarer type of cancer, it is one of the most fatal once diagnosed. For reference, pancreatic cancer comprises 3% of all cancer malignancies; however, it is responsible for nearly 10% of all cancer-related deaths.

Of the two categories of pancreatic cancer, 95% are exocrine pancreatic cancers, which impact the cells that produce enzymes, while endocrine cancers only account for 5% of pancreatic cancer cases.

An article in Cancers suggests that pancreatic ductal adenocarcinoma (PDAC) is the most common type of pancreatic cancer. The average 5-year survival rate of PDAC is as low as 12% since most patients are not diagnosed until the advanced stages of the disease.

Risk factors

Although some factors, including germline mutations, mucinous cystic lesions, or chronic pancreatitis, contribute to 10–15% of pancreatic ductal adenocarcinoma (PDAC) cases, researchers in the Cell article note that most cases cannot be attributed to one risk factor or condition.

Peutz–Jeghers syndrome, hereditary breast or ovarian cancer syndrome, familial atypical melanoma, ataxia telangiectasia syndrome, and Lynch syndrome are genetic risk factors for PDAC; however, they only account for 5–10% of all cases.

Generally, environmental factors are a better indicator of PDAC risk. Heavy alcohol consumption, tobacco use, type 2 diabetes, obesity, and chronic pancreatitis correlate with an elevated risk of PDAC.

For example, the National Cancer Institute estimates that 1% of people with new-onset diabetes are diagnosed with pancreatic cancer within three years of their diabetes diagnosis. Additionally, 25% of patients diagnosed with pancreatic cancer have diabetes.

Symptoms

In the early disease stages, patients with PDAC have symptoms including epigastric or back pain, nausea, bloating, distention, or changes in stool. However, since these symptoms are nonspecific and can be mistaken for other, less fatal conditions, most patients are only diagnosed once the disease has advanced past resection potential.

While patients with advanced PDAC may continue to present with nonspecific symptoms such as loss of appetite and weight loss; others also have more severe symptoms such as obstructive jaundice and new or exacerbated diabetes.

Surgical tumor resection and chemotherapy are the standard treatment options for patients with PDAC. However, few patients, 15–20%, are diagnosed early enough for surgical resection to be a viable option. In the remaining 80–85% of patients, the cancer has already metastasized beyond surgical intervention when they are diagnosed.

Furthermore, approximately 75% of patients who undergo surgery will have a recurrence within two years.

Pancreatic Cancer Screenings

There are multiple existing diagnostic tools for detecting pancreatic cancer. However, most of these diagnostic tools are ineffective at early diagnosis because they are not initiated until later disease stages. Typically, pancreatic cancer diagnosis includes one or more of the following tools:

• Medical imaging, including computed tomography (CT) or magnetic resonance imaging (MRI)
• Endoscopic ultrasounds
• Blood tests
• Biopsies

The most widely used pancreatic cancer diagnostic tool is a multidetector CT angiography with a dual phasic pancreatic protocol.

The Biomedicines article notes that a dedicated CT protocol for pancreatic lesions has been developed, citing a 90% specificity and sensitivity. “This protocol provides a standardized template for reporting that specifies the terminology to be used for solid pancreatic neoplasms, with the aim of improving the patient-management decision-making process and optimizing treatment recommendations,” added researchers.

Beyond detecting lesions in late stages, CT has also been proven to detect local atrophy or fatty metamorphosis of the pancreatic parenchyma, which lends itself to earlier disease detection.

MRI is thought to be as specific and sensitive as CT in detecting pancreatic cancer through images of morphological changes; however, since MRIs are more expensive and less accessible, most facilities opt for CT scans.

In addition to noninvasive imaging, endoscopic ultrasounds provide the most detailed image of the pancreas, meaning it is more likely to detect disease at the early stages.

According to Biomedicines, existing diagnostic tools for pancreatic cancer are “highly sensitive and specific for identifying lesions within the pancreas;” however, applying these tools in early disease stages remains a problem in the field.

Recent Advancements and Gaps

Researchers in Biomedicines highlight multiple advancements to current diagnostic approaches that could revolutionize the diagnostic pathway.

Machine Learning Integrations

For example, one of the primary limitations of imaging diagnostics and cytology is the need for experienced professionals — typically radiologists and pathologists — to interpret findings. While some facilities may readily access this talent, others do not. One strategy being explored to address these concerns is applying machine learning (ML) or deep learning (DL) algorithms that can detect features of pancreatic cancer and incorporate them into the diagnostic pipeline.

For example, in a 2022 study published in Clinical Biomarkers, researchers developed the naïve Bayes classifier for pancreatic cancer detection using 108 retrospective CT scans. The system could detect pancreatic cancer with an 86% accuracy rate.

Similar models and systems have been developed; however, more research is needed to validate the models and support clinical applications.

“Machine learning has the potential to dramatically change how pancreatic cancer is diagnosed. Moreover, if the cost and turnaround times for processing CT data can be controlled, imaging-based screening could potentially be used to assess larger asymptomatic populations. However, machine learning has potential drawbacks that need to be carefully considered,” emphasized researchers in the Biomedicines article.

Serum Biomarker Tests

In an attempt to search for lower-cost diagnostic approaches, some clinicians and investigators are looking at the utility of serum biomarker tests. Serum biomarker tests are low-cost and easy to apply.

Most new serum biomarker screens have combined new biomarkers, such as cell-free DNA, cell-free RNA, cell-free proteins, circulating tumor cells, and exosomes, and existing biomarkers, such as CA19-9 and CEA. Elevated levels of these biomarkers have been linked to an increased risk of pancreatic cancer; however, multiple studies suggest that they are not reliable as independent tests.

More recently, CA19-9 tests have been combined with serum tests for endostatin and collagen IV, circulating tumor-associated autoantibodies, methylated HOXA1, methylated SST, pancreatic elastase-1, pancreatic amylase, and pancreatic lipase.

Researchers have also used serum biomarker tests that evaluate and detect mutated mitochondrial DNA in exosome samples to detect pancreatic cancer.

Biopsy

Beyond new diagnostic tools, clinicians and researchers are working on fine-tuning the biopsy process by identifying optimal techniques, needle types, new methods, or contrast agents.

Screening

While precise and validated diagnostic tools are critical to early disease diagnosis, researchers also look for ways to improve surveillance. Diagnostic tools are only effective if initiated. The Biomedicines article points out that, theoretically, universal screening for pancreatic cancer could improve early disease detection and survival rates, but currently available tests make that route a financially unattainable goal.

“Another approach would be to screen only those individuals at high risk of developing this malignancy, although attempts to identify risk factors associated with pancreatic cancer have met with limited success,” added researchers in the Biomedicines article. However, it is difficult to identify a single risk factor that predicts pancreatic cancer.

Currently, the NCI is funding a study called the New Onset Diabetes (NOD) that is enrolling 10,000 patients with new onset diabetes or prediabetes to develop a blood test for patients newly diagnosed with diabetes who may be at risk for pancreatic cancer.

Additionally, the Pancreatic Cancer Detection Consortium includes NCI teams working on developing a blood test to screen for pancreatic cancer in the general population.

Although available screening tools can detect pancreatic cancer well, researchers and public health experts need to collaborate on ways to get suitable patients screened before the disease progresses.