Endoscopic ultrasound in chronic pancreatitis

INTRODUCTION — Endoscopic ultrasound (EUS) was initially developed to improve imaging of the pancreas [1-4]. The close proximity of the pancreas to the gastric and duodenal lumen permits EUS to obtain high-resolution imaging without interference from overlying bowel gas.

The EUS diagnosis of chronic pancreatitis (CP) relies on quantitative and qualitative parenchymal and ductal criteria of which several have been published. It is generally accepted that, in the absence of any criteria, CP is unlikely, whereas in the presence of ≥5 criteria, CP is likely even though endoscopic retrograde cholangiography (ERCP) and standard tests of pancreatic function may still be normal. The clinical significance of fewer (one to four) features found on EUS is unclear, particularly when other diagnostic tests such as ERCP and function testing are normal.

EUS has allowed for the recognition of several features of CP that had not been appreciated with other imaging modalities. These include hyperechoic margins of the pancreatic duct, subtle lobularity of the parenchyma, small cystic changes in the parenchyma, and side branch duct ectasia (image 1A-B). The ability to detect these changes has raised questions as to how chronic pancreatitis should be defined and whether EUS may be overly sensitive. However, accumulating evidence suggests that these early changes detected by EUS correlate with histologic changes of CP and may predict progression to more overt disease.

NORMAL PANCREATIC EUS — Several normal sonographic findings of the pancreas have been described, which allow for appreciation of the abnormalities that can be seen on EUS. The following findings were noted in a study involving 130,951 patients who underwent a screening examination with transabdominal ultrasonography (TUS) using a 3.5 or 5 mHz probe [5]:

●Pancreatic duct diameter, which is measured similarly by TUS and EUS, was dilated (>3 mm) in only 0.49 percent of individuals and was more common in men and in older individuals. There was a strong trend toward increasing duct diameter with age.

●Cystic lesions were detected in 0.21 percent and calcifications in 0.05 percent of individuals. However, TUS may underestimate the prevalence of these abnormalities compared with EUS.

Several studies have evaluated the pancreas with EUS in populations of patients without clinical features or obvious risk factors for chronic pancreatitis such as those undergoing nonpancreatic tumor staging, evaluation of submucosal tumors or portal hypertension [6-9]. However, whether these studies truly reflected findings in a population of patients with a "normal" pancreas is unclear. Some patients may have had risk factors for chronic pancreatitis (such as alcohol use disorder) while others with nonpancreatic malignancies had severe cachexia, which can alter the sonographic appearance of the pancreas.

●In a report of 20 patients, the pancreatic parenchyma was described as having a homogeneous, fine, granular appearance with smooth margins [8]. The pancreatic duct diameter was 1.9 mm on average (range 1.5 to 2.4 mm).

●In another series of 25 patients, the pancreatic parenchyma was described as being homogeneous and finely reticulated without evidence of side branch ectasia [7]. A ventral anlage (echogenic difference between the ventral and dorsal pancreas) was seen in 68 percent of patients. No cysts or stones were described. The main pancreatic duct was uniformly tubular in shape with anechoic walls and a mean diameter (in the pancreatic body at the portal vein confluence) of 1.7 mm (range 1 to 3 mm). Only two patients had a duct diameter of 3 mm. Side branches were visible in 8 of 25 patients.

●A third report focused on a group of healthy volunteers without a prior history of abdominal pain or alcohol use disorder [10]. The pancreatic parenchyma was described as being uniform and more echogenic than liver. A ventral anlage was detected in 45 percent. No cysts were seen. The main pancreatic duct diameter was 2.4 mm (range 0.8 to 3.6) in the head, 1.8 mm (0.9 to 3.0) in the body, and 1.2 mm (range, 0.5 to 2.0) in the tail. Side branches were visible but narrow (mean diameter 0.8 mm, head; 0.5 mm, body; 0.3 mm, tail).

REFERENCE STANDARDS — There is no agreement on an appropriate reference standard for chronic pancreatitis (CP), particularly early or minimal change CP [11]. Nevertheless, findings on endoscopic ultrasound (EUS) have been correlated with other measures of CP including histology, pancreatic function testing, and other forms of radiologic imaging.

Histology — One of the most obvious candidates for a reference standard is histology. However, there is little agreement on the histologic features of CP, and features that have been described are often focally distributed or subtle [12]. Despite these problems, several studies have compared results of EUS with subsequent resection specimens. These studies suggest that EUS detects ductal and parenchymal changes in patients with suspected CP.

●One study focused on 34 patients who underwent EUS followed by pancreatectomy or open surgical biopsy (21 patients for CP, 12 patients for pancreatic cancer) [13]. Overall, 68 percent were considered to have CP based upon the histologic findings. The total number of EUS criteria present was predictive of histologic CP. The sensitivity and specificity using a threshold diagnosis for ≥3 criteria were 87 percent and 64 percent, for ≥4 criteria 78 percent and 73 percent, for ≥5 criteria 60 percent and 83 percent, and for ≥6 criteria 43 percent and 91 percent. The authors concluded that a minimum of four or more criteria was optimal.

●Two reports evaluated EUS in patients with a heterotopic pancreas [14,15]. Findings confirmed pathologically included small cystic changes, hyperechoic foci (corresponding to adipose tissue), and small duct dilation.

●Two studies evaluated intraductal ultrasound (IDUS) in freshly excised or snap-frozen pancreas specimens. One report compared IDUS with histology in seven "normal" pancreas sections post mortem, and one Whipple resection [16]. IDUS criteria for hyperechoic strands corresponded to fibrous bands around cystic structures. Another study compared IDUS with histology in freshly excised pancreatic tissue in 11 patients with CP [17]. IDUS detected lobularity (which corresponded to interlobular fibrosis) and hyperechoic duct margins (which corresponded to periductal fibrosis). All benign areas had preservation of a fine reticular pattern.

●A study of 27 patients with CP compared the results of EUS with the findings on fine-needle aspiration biopsy (FNA) and endoscopic retrograde cholangiopancreatography (ERCP) [18]. EUS findings correlated with the severity of CP according to the Cambridge classification at ERCP in 5 of 8 patients with grade I, 11 of 13 patients with grade II, and 10 of 10 patients with grade III disease. Furthermore, EUS-FNA increased the negative predictive value to 100 percent and the specificity to 64 percent.

Pancreatography — Multiple studies have compared the results of EUS with ERCP in patients with abdominal pain and suspected CP [7,8,10,19-21]. Four of these [7,10,20,22] used a standardized EUS and ERCP grading system (the Cambridge classification), making them directly comparable. Each evaluated the pancreas for the presence of 9 to 11 described criteria and then considered EUS to be abnormal if the total number of criteria exceeded a threshold number (eg, ≥3, ≥4, or ≥5). EUS and ERCP agreed in approximately 80 percent of cases when the threshold of ≥3 criteria was considered. In cases where they disagreed, the majority (74 percent) are abnormal by EUS and normal by ERCP. It is unclear if EUS is more sensitive to mild changes of CP than ERCP or, again, if EUS was "overdiagnosing" early CP.

Functional testing — At least three studies with a total of 179 patients have compared EUS with pancreatic function studies in which pancreatic juice was analyzed following stimulation with secretin [7,10,21]. (See "Approach to the adult patient with suspected malabsorption".).

Overall, EUS and functional testing agreed in 75 percent of cases. Similar to the results of pancreatography, the main source of disagreement (71 percent) occurred in patients with an abnormal EUS but a normal functional test. Both EUS and ERCP also appear to correlate well with fecal elastase, a noninvasive measure of pancreatic exocrine function [23].

Natural history — Natural history may be the most clinically relevant standard for early CP. A diagnosis of "mild" CP based upon EUS, which then progresses to more severe CP as diagnosed by other tests (EUS-positive, ERCP-positive functional test), is likely to be a correct diagnosis. Unfortunately, there are only limited data on the long-term natural history of "mild" CP diagnosed by EUS.

●A cross-sectional study of patients with and without abdominal pain showed that EUS diagnosis of CP (≥4 criteria) was positive in 89 percent of patients with alcohol use disorder with abdominal pain, but also in 58 percent of patients with alcohol use disorder but without pain and 0 percent of control patients (no alcohol use, no abdominal pain) [24].

●Another report described the progression of pancreatic disease in 17 asymptomatic individuals with alcohol use disorder with an abnormal EUS but a normal ERCP [21]. Follow-up EUS examinations for 12 to 38 months (mean 22 months) did not identify progression to more overt (ERCP-positive) disease.

The follow-up period was notably short considering the natural history of chronic pancreatitis, which typically progresses over many years. Thus, it is not surprising that ductal changes were not detected. On the other hand, another study suggested that ductal changes may be detectable even with such short follow-up. The study focused on 38 patients with a history of chronic alcohol use and recurrent abdominal pain [25]. At the time of enrollment, 32 of 38 patients had an abnormal EUS but a normal ERCP. After a median follow-up of 18 months (range 6 to 25 months), 22 of 32 patients developed changes of CP at ERCP (12 patients grade I and 10 patients grade II according to the Cambridge classification).

The only significant differences between these two studies appeared to be related to inclusion criteria: patients with alcohol-associated cirrhosis were included in the first study [21], while the second study focused on patients with abdominal pain and recurrent pancreatitis [25]. Further long-term follow-up data are needed to help clarify these relationships.

Response to therapy — One study focused on 43 patients who had characteristic symptoms of CP, but who had normal or equivocal findings on ERCP, CT or TUS [12]. Patients were treated medically (with pancreatic enzyme replacement therapy [PERT], a low-fat diet, and at least three trials of bowel rest with total parenteral nutrition).

A pancreatic resection was performed in 16 nonresponding patients. The histologic appearance in the pancreas of these patients showed subtle but distinct evidence of minimal-change chronic pancreatitis. These changes were distributed focally throughout the gland and included lymphocytic cell infiltrates, intralobular and periductal fibrosis, and focal ductal dilation with inspissated protein plugs. Nine of the 16 patients had complete or significant improvement in pain after total pancreatectomy, whereas five did not respond and one died of unrelated causes. The histologic changes and response to pancreatectomy suggest that these patients had CP despite normal imaging and functional testing. However, a placebo response cannot be excluded. Furthermore, unblinded interpretation of the resection specimens may have also introduced bias.

INTEROBSERVER RELIABILITY — The interobserver reliability of endoscopic ultrasound (EUS) in the evaluation of chronic pancreatitis (CP) varies depending upon the specific finding. In a study comparing interpretation among three experienced endosonographers, agreement was 88 percent for hyperechoic foci, 94 percent for focal reduced echogenicity, 94 percent for lobularity, 83 percent for hyperechoic duct margins, and 94 percent for duct irregularity [10].

A similarly designed study compared the interpretations of 11 highly experienced endosonographers [26]. There was moderate overall agreement for the final diagnosis of CP (kappa [k] = 0.45; a perfect test would have a k = 1.00) and for individual criteria of duct dilation (k = 0.60) and lobularity (k = 0.51), but poor for seven other criteria (k <0.4). The expert panel had consensus or near consensus agreement (>90 percent) on 206 of 450 (46 percent) individual examination criteria including 22 of 45 diagnoses of CP [26]. To further improve the reliability, an international working group published a set of minimum standard terminology, including definitions, for many of the EUS criteria of CP (table 1) [26].

These estimates of reliability are comparable to other commonly used endoscopic procedures such as bleeding ulcer stigmata (k = 0.34 to 0.66) [27] and certain radiologic procedures such as brain CT for stroke localization (k = 0.56 to 0.62) [28], and better than the physical diagnosis of heart sounds (k = 0.05 to 0.18) [29]. However, reliability in practice is often lower than that observed in controlled trials.

LINEAR VERSUS RADIAL EUS — Endoscopic ultrasound can be performed with either a linear or a radial echoendoscope. (See "Endoscopic ultrasound-guided fine needle aspiration in the gastrointestinal tract", section on 'Equipment'.)

Endoscopic ultrasound for evaluating chronic pancreatitis is most often performed using linear EUS. However, in a noninferiority study of 100 patients, there was no significant difference between linear EUS and radial EUS with regard to sensitivity or specificity (44 versus 68 percent and 95 versus 95 percent, respectively) [30]. Accuracy was also similar in the two modalities (74 percent for linear EUS and 84 percent for radial EUS).

While radial EUS was not inferior to linear EUS in the study mentioned above, there are potential advantages to linear EUS over radial EUS. Linear EUS has the ability to perform fine-needle aspiration and/or biopsy, which is advantageous if lesions of concern are identified and need to biopsied.

THRESHOLD TO DIAGNOSE OR EXCLUDE CP BY EUS — As noted above, there are many sonographic features that may indicate the presence of CP. The threshold for diagnosing CP based on EUS can be varied (eg, ≥3, ≥4, or ≥5 criteria).

Conventional criteria (Minimal Standard Terminology) for diagnosing CP based on EUS include [31]:

●Parenchyma abnormalities

•Hyperechoic foci

•Hyperechoic strands

•Lobular contour

•Cysts

●Ductal abnormalities

•Main duct dilation

•Duct irregularity

•Hyperechoic margins

•Visible side branches

•Stones

The sensitivity and specificity of EUS compared with other reference standards depends upon which threshold is chosen. (See "Glossary of common biostatistical and epidemiological terms".)

●If a low threshold is used (>1 to 2 criteria) the sensitivity (and negative predictive value [NPV]) will be high, but the specificity (and positive predictive value [PPV]) will be low.

●If a higher threshold is used (>5 to 6 criteria), the sensitivity (and NPV) will be low, but the specificity (and PPV) will be high.

The optimal cutoff depends upon the purpose of the examination:

●If the purpose is to exclude disease, a low threshold will give the best negative predictive value. For example, a patient with only zero to one criteria of CP by EUS has a >90 percent chance of having a normal endoscopic retrograde cholangiopancreatography (ERCP) [20] (and presumably does not have CP). In such a case, further testing is unlikely to identify CP.

●If the purpose is to establish the diagnosis with a high degree of certainty, a high threshold such as ≥6 EUS criteria gives a positive predictive value of >80 percent of having an abnormal ERCP. Those with fewer criteria may require further evidence to establish the diagnosis of CP.

Two other factors must be taken into account when diagnosing CP based upon EUS criteria. All criteria may not be equally important. As an example, the presence of intraductal calcifications (stones) alone is highly suggestive of CP even in the absence of other criteria. In addition, there are age-related changes in the pancreas that may affect the diagnostic threshold [32]. As noted above, the pancreatic duct becomes progressively wider as individuals age [5,33]. A 4 mm main pancreatic duct may be normal for a 70-year-old person, but abnormal for a 30 year old. There is no accepted scoring system that factors in these effects.

One approach is to require a higher threshold (eg, ≥5 criteria) for older individuals and a lower threshold (eg, ≥4 criteria) for a younger individual. A lower threshold was suggested in a study of 120 patients (ages 40 to 61 years) without known pancreatic disease; the authors suggested that age alone was unlikely to be responsible for sonographic changes when at least three criteria suggesting chronic pancreatitis were present [33]. The most common sonographic abnormality in all age groups was hyperechoic stranding. At least one parenchymal and/or ductular abnormality was identified in 28 percent of patients.

Rosemont classification — An international consensus panel (convened in Rosemont, Illinois) developed consensus criteria for EUS features of chronic pancreatitis [34]. Major criteria included: (1) hyperechoic foci with shadowing and main pancreatic duct calculi, (2) lobularity with honeycombing. Minor criteria included cysts, dilated ducts ≥3.5 mm, irregular pancreatic duct contour, dilated side branches ≥1 mm, hyperechoic duct wall, strands, nonshadowing hyperechoic foci, and lobularity with noncontiguous lobules. The group proposed a classification scheme based on the combinations of these criteria. Its validity requires additional studies in which the classifications are compared with clinically relevant reference standards.

SUPPLEMENTAL EUS TECHNIQUES — Several variations of standard EUS techniques have been developed to improve the diagnosis of CP, but none is used widely [35].

●Intraductal ultrasonography has been useful for discriminating pancreatic duct strictures due to CP from those due to malignancy. (See "Intraductal ultrasound for evaluating the pancreaticobiliary ductal system".)

●Contrast agents such as Albuminex may help distinguish between pancreatic cancer and CP since the former tends to be hypovascular [36].

●Secretin stimulation enhances pancreatic duct dilation in patients with CP and has long been used in transabdominal ultrasonography, where its benefit remains to be clearly established. Its role during EUS (secretin-stimulated EUS or SSEUS) was evaluated in a study that included 120 patients with suspected pancreatic disease who were compared with 20 controls [6]. An increase in the main pancreatic duct of ≥1 mm that persisted for at least 15 minutes was considered abnormal. SSEUS was compared with a diagnosis of CP based upon an abnormal pancreatogram, calcifications, or bicarbonate secretion.

The SSEUS had a sensitivity of 80 percent and specificity of 96 percent for the diagnosis of CP when considering the presence of any one of these reference standards. SSEUS also predicted which patients with pancreas divisum would respond to stenting; an abnormal SSEUS had a positive predictive value of 93 percent and a negative predictive value of 62 percent for response to stent therapy. Although promising, these results await confirmation by other studies.

DISCRIMINATION FROM PANCREATIC CANCER — Chronic pancreatitis shares clinical and radiologic features with pancreatic cancer. As a result, there may be diagnostic confusion, which can lead to unnecessary resection in those with CP. The advent of EUS-guided sampling with fine needle aspiration and/or biopsy has largely overcome this issue because sampling may provide a tissue diagnosis (see "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Diagnostic approach'):  

●EUS fine-needle aspiration (FNA) has been used to sample mucosal and submucosal lesions in which prior conventional endoscopic biopsies have been nondiagnostic. The procedure is most commonly used to sample peri-intestinal structures such as lymph nodes and masses in the pancreas, liver, adrenal gland, and bile duct. Studies suggest that the sensitivity of EUS-FNA for diagnosis of CP may not be sufficiently high to establish the diagnosis, but high specificity may be helpful for exclusion of other diagnoses. (See "Endoscopic ultrasound-guided fine needle aspiration in the gastrointestinal tract".)

●In some patients, EUS findings may be unrevealing, despite clinical suspicion for pancreatic cancer. One study focused on a group of 80 such patients in whom pancreatic cancer was suspected, either because of elevated serum concentrations of CA 19-9 without other definitive evidence of pancreatic cancer, subtle abnormalities on CT, or unexplained abdominal pain and/or weight loss [37]. Follow-up was available for at least six months in 76 patients (95 percent) after EUS. No patients with a normal EUS developed pancreatic cancer. One patient who was diagnosed with chronic pancreatitis by EUS was subsequently found to have pancreatic cancer at surgery. The authors concluded that pancreatic cancer is unlikely in patients with equivocal clinical suspicion for pancreatic cancer who have a normal pancreatic EUS.

●Another study compared EUS findings in 21 patients with pancreatic cancer with 14 patients with focal pancreatitis [38]. The final diagnosis was proven histologically in all patients. The sensitivity, specificity, and accuracy of the endosonographers' impression at the time of the EUS examination were 89, 79, and 85 percent, respectively. After a blind interpretation of the videotapes, the sensitivity, specificity, and accuracy were 73, 100, and 83 percent, respectively.

These studies underscore the limitations of diagnostic EUS in discriminating CP from tumors. Cytopathologists face equivalent challenges when trying to discriminate inflammatory from neoplastic cells. In the future, translational research with application of specific molecular (or tumor) markers to FNA specimens may provide a significant aid in discriminating CP from tumors.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Chronic pancreatitis and pancreatic exocrine insufficiency".)

SUMMARY AND RECOMMENDATIONS

●Endoscopic ultrasound (EUS) detects changes of mild chronic pancreatitis (CP) that may not be detectable with other imaging modalities or functional testing but can be confirmed by histology. However, controversy remains regarding the diagnosis of "early" chronic pancreatitis based upon EUS changes alone. (See 'Reference standards' above.)

●There are multiple sonographic features that may indicate the presence of chronic pancreatitis. The threshold for diagnosing chronic pancreatitis based on EUS can be varied (eg, ≥3, ≥4, or ≥5 criteria). Criteria for diagnosing chronic pancreatitis based on EUS include (see 'Threshold to diagnose or exclude CP by EUS' above):

Parenchyma abnormalities:

•Hyperechoic foci

•Hyperechoic strands

•Lobular contour

•Cysts

Ductal abnormalities:

•Main duct dilation

•Duct irregularity

•Hyperechoic margins

•Visible side branches

•Stones

●Chronic pancreatitis shares clinical and radiologic features with pancreatic cancer. As a result, there may be diagnostic confusion, which can lead to unnecessary resection in those with chronic pancreatitis. The advent of EUS-guided sampling with fine needle aspiration and/or biopsy has largely overcome this issue because sampling may provide a tissue diagnosis. (See 'Discrimination from pancreatic cancer' above and "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Diagnostic approach'.)