Clinical Trial Details
— Status: Completed
Administrative data
NCT number |
NCT06278272 |
Other study ID # |
AIPEI |
Secondary ID |
|
Status |
Completed |
Phase |
|
First received |
|
Last updated |
|
Start date |
March 1, 2023 |
Est. completion date |
January 1, 2024 |
Study information
Verified date |
January 2024 |
Source |
Changhai Hospital |
Contact |
n/a |
Is FDA regulated |
No |
Health authority |
|
Study type |
Observational
|
Clinical Trial Summary
Early assessment of pancreatic exocrine insufficiency (PEI) is crucial for determining
appropriate chronic pancreatitis (CP) treatment plans, thereby avoiding unnecessary suffering
and further complications in patients. A total of 504 patients with CP who underwent fecal
elastase-1 test and contrast-enhanced CT at Changhai Hospital between January 2018 and April
2023 were enrolled in this study. The investigators aim to establish a fully automated
workflow to establish a PEI classification model based on radiomic features, semantic
features and deep learning features on enhanced CT images for evaluating the severity of PEI.
Description:
Chronic pancreatitis (CP) is a chronic inflammatory disease characterized by abdominal pain,
recurrent inflammation, and pancreatic fibrosis. The global incidence of CP is about
10/100,000 and shows an increasing trend year by year.
Pancreatic exocrine insufficiency (PEI) is a significant functional alteration in chronic
pancreatitis. The prevalence of PEI in CP patients within 10 to 15 years after diagnosis is
about 35% to 50%, and the prevalence increases significantly 15 years after
diagnosis.Clinical manifestations of PEI vary among individuals and may include symptoms such
as diarrhea, weight loss, abdominal pain, bloating, and steatorrhea. Some patients may remain
asymptomatic, leading to deficiencies in fat-soluble vitamins and micronutrients due to
abnormal digestion of macronutrients, a condition referred to as "subclinical PEI".Due to
nutritional deficiencies, PEI patients are at an increased risk of complications related to
malnutrition, osteoporosis, cardiovascular diseases, and elevated mortality rates.
Importantly, the quality of life for PEI patients is significantly reduced, and timely
diagnosis and treatment are crucial to prevent severe consequences. Early and appropriate use
of pancreatic enzyme replacement therapy can enhance patient well-being and improve overall
quality of life. Therefore, regular screening for PEI and prompt treatment are essential for
CP patients to reduce the risk of complications and improve prognosis.It is recommended by
current clinical guidelines that pancreatic exocrine function should be assessed at the time
of CP diagnosis, and if no PEI is detected, annual screening is advised. Despite these
recommendations, PEI is still frequently misdiagnosed and inadequately treated in clinical
practice.
Pancreatic exocrine function testing involves both direct and indirect methods. Direct tests,
such as the cholecystokinin-pancreozymin stimulation test, are considered the most sensitive
and specific means of assessing pancreatic exocrine function. This test requires intravenous
infusion or injection of cholecystokinin to stimulate pancreatic secretion, followed by
duodenal content collection via a catheter for the measurement of pancreatic fluid secretion.
However, due to its high cost, invasiveness, and significant discomfort for patients, this
test is seldom used in clinical practice.Indirect tests encompass fecal tests, breath tests,
urine tests, and blood tests. Among these, fecal elastase-1 (Fe-1) detection is a currently
stable, accurate, and convenient indirect method. When Fe-1 levels range between 100-200
μg/g, it suggests mild to moderate PEI, while Fe-1 levels below 100 μg/g indicate severe PEI.
Several studies have demonstrated that a Fe-1 level below 200 μg/g has a specificity greater
than 90% for diagnosing PEI. However, the limited availability of the assay kit has
restricted its widespread use.
In imaging examinations, the use of secretin-enhanced magnetic resonance
cholangiopancreatography to measure pancreatic fluid flow after secretin stimulation is a
non-invasive method for assessing PEI. However, its clinical feasibility is limited due to
its time-consuming nature, high cost, complexity, and the lack of widespread availability of
the required reagents. Computed Tomography (CT) is the most widely used imaging examination
in the clinical assessment of CP, but there is no evidence supporting its utility in
detecting PEI. Therefore, there is a need for a new, non-invasive, and convenient method for
early detection of PEI.
In recent years, deep learning has shown significant promise in medical image analysis.
Compared to traditional machine learning, deep learning utilizes models to analyze data and
extract higher-level features. It then uses these features to derive results, enabling the
modeling of complex relationships within the data. Commonly used methods include
convolutional neural networks, generative adversarial models, and others. Currently, deep
learning has been widely applied in pancreatic diseases, including tumor detection,
differential diagnosis of pancreatic cancer, and predicting patient prognosis. However, there
has been no research utilizing deep learning to assess PEI.This study aims to develop and
validate a deep learning model based on fully automated pancreatic segmentation for
evaluating pancreatic exocrine function. Additionally, The investigators will collaborate
with radiologists to jointly assess the model's performance.