The importance of pancreatic inflammation in endosonographic diagnostics of solid pancreatic masses

(1)

DOI:

Original papers

The importance of pancreatic inflammation in endosonographic diagnostics of solid pancreatic masses

Francesco Vitali

¹

, Deike Strobel

¹

, Luca Frulloni

²

, Marc Heinrich

³

, Lukas Pfeifer

¹

, Ruediger S. Goertz

¹

, Gheorghe Hundorfean

¹

, David L. Wachter

⁴

, Robert Gruetzmann

⁵

, Thomas Bernatik

⁶

, Markus F. Neurath

¹

, Dane Wildner

¹

1Department of Internal Medicine 1, Friedrich-Alexander-University, Erlangen- Nuremberg, Germany, ²Department of Medicine, Pancreas Center, University of Verona, Verona, Italy, ³Department of Radiology, Friedrich-Alexander- University, Erlangen-Nuremberg, Germany, ⁴Institute of Pathology, Friedrich-Alexander-University,

Erlangen-Nuremberg, Germany, ⁵Department of General and Abdominal Surgery, Friedrich-Alexander-University, Erlangen-Nuremberg, Germany, ⁶Department of Internal Medicine, Kreisklinik Ebersberg, Germany

Received 07.07.2018 Accepted 20.10.2018 Med Ultrason

2018, Vol. 20, No 4, 427-435

Corresponding author: Francesco Vitali, M.D.

Department of Internal Medicine 1, Friedrich- Alexander-University, Erlangen-Nuernberg, Ulmenweg 18, 91054 Erlangen, Germany Phone: ++ 49/91 31/8 53 50 00

Fax: ++ 49/91 31/8 53 52 52

E-mail: [email protected]

Introduction

Endosonography (EUS) is an established technique for the clinical work-up of a newly recognized pancreatic

mass. Reported sensitivities range from 93 to 98% for diagnosing pancreatic cancer, especially for lesions smaller than 3.0 cm [1-3]. Cross sectional imaging like computed tomography (CT) and magnetic resonance imaging (MRI) sometimes show abnormal findings that raise the suspicion of pancreatic cancer, resulting in a further work-up [4]. Despite advancements in cross sectional imaging and the use of EUS, published retrospective surgical studies demonstrate that the incidence of pancreatic resections performed for benign lesions mimicking pancreatic malignancy did not change from 1983 to 2014 and is still about 9-10% for pancreatic head lesions [5-19].

New diagnostic modalities such as contrast enhanced-EUS, quantitative contrast enhanced analysis, Abstract

Aims: Endosonography (EUS) is one of the main diagnostic tools for the differential diagnosis of pancreatic masses. The aim of our study was to describe the value of this technique in the work-up of solid pancreatic lesions, considering the influence of the morphological evidence of pancreatic inflammation in the diagnostic process. Material and methods: Retrospec- tive analysis of prospectively collected data in our tertiary University center. From March 2007 to October 2015, 218 patients underwent EUS for a suspected solid pancreatic neoplasm (based on previous cross-sectional imaging results, idiopatic acute pancreatitis, weight loss, pancreatic hyperenzymemia, painless jaundice or elevated Ca 19-9 values). Results: Malignant lesions were diagnosed in 98 (45%) patients. Sensitivity of EUS for malignancy was 91% and specificity 89.2%. Signs of pancreatic inflammation in the surrounding pancreatic parenchyma around the focal lesion were present in 97 patients (44.4%) (more often in men, smokers and drinkers, and the most common etiology was focal chronic pancreatitis) and in these patients the sensitivity and sensibility dropped to 44% and 87.1%, respectively. In patients without signs of pancreatic inflammation, the pancreatic focal lesions were adenocarcinoma, neuroendocrine tumor, ventral/dorsal split, non-pancreatic pathology, pancreatic lipomatosis and autoimmune pancreatitis. Conclusion: Pancreatic inflammation (either focal or involving the whole gland) lowers the diagnostic sensibility of EUS in the work- up of pancreatic masses suspected for cancer, requiring further invasive diagnostic methods. Focal autoimmune pancreatitis and paraduodenal pancreatitis are still confused with pancreatic cancer, even in the absence of pancreatic inflammation.

Keywords: endoscopic ultrasound; pancreatic cancer; benign pancreatic lesion; pancreatitis

DOI: 10.11152/mu-1641

(2)

fine needle aspiration (FNA) and core biopsy have been added to improve diagnostic performance [20,21]. Re- cently, real-time elastography and three-dimensional reconstruction have become available, enabling differentiation of typical ductal pancreatic adenocarcinoma from other neoplasms and benign solid lesions with an accuracy of approximately 90% [22-26]. However, despite all these advanced diagnostic tools, the diagnosis is still challenging and radical surgery is still performed [6].

Retrospective data supports evidence that a suspicious pancreatic mass detected upon contrast-enhanced CT should directly undergo surgery without further investigation [27]. Consensus among surgeons recom- mends primary surgery for pancreatic masses, except in cases highly suspicious of autoimmune pancreatitis (AIP) [28]. On the other hand, without preoperative diagnosis, an unacceptable amount of patients might unnecessarily be exposed to radical pancreatic surgery with consider- able morbidity and mortality for pre-malignant or benign lesions [3]. This situation is reported to be in Germany up to 11.1% in very high volume hospitals to 20.2% in very low volume hospitals [29].

AIP, which is a typical mimicry of pancreatic cancer, is increasingly being diagnosed and treated worldwide [30]. In 2011, the International Consensus Diagnostic Cri- teria (ICDC) for diagnosing AIP was published. Paradu- odenal pancreatitis may also present as a mass-forming process and it is increasingly being diagnosed since the publication of histopathological diagnostic criteria [31] and defined clinical and imaging features [32-35].

The aim of this study was to determine the diagnostic sensitivity and specificity of EUS for pancreatic lesions suspicious for malignancy. In addition, we investigated the influence of acute and chronic inflammatory changes of the pancreas on the diagnostic presentation and on the final diagnoses.

Material and methods

We conducted a retrospective analysis of prospectively collected data in our tertiary care university hospital. The present study enrolled patients with focal solid pancreatic changes referred from gastroenterological and radiological practices, general practitioners and primary and secondary care centers of our region. Complex cases were mainly referred to our surgical department and discussed in our interdisciplinary board. Some patients had primarily been examined in other departments and were admitted to our center for a second opinion or for a further workup. From March 2007 to October 2015, 2356 EUS examinations were performed. Indications for examination were screened from examining patients’ clini-

cal files in order to select those who underwent EUS for a suspected pancreatic solid lesion (fig 1). EUS was performed by 3 experienced endosonographers (each more than 1000 procedures) of our institution, using Pentax EG-3630UR Radial and EG-3830UT echoendoscopes (PENTAX Europe GmbH, Hamburg) with a Hitachi EUB 6500 ultrasound system (Hitachi Medical Systems GmbH, Wiesbaden). The equipment did not change during the study period. EUS was performed with radial- scanning echoendoscopes. Curvilinear echoendoscopes were used for FNA. Standard EUS for the evaluation of the pancreas was performed. If the EUS findings raised a suspicion that a focal lesion was present, histological specimens were obtained, at the discretion of the endo- sonographer, with EUS-FNA using a curvilinear array echoendoscope. The needle used was a 19 G needle to achieve histological evaluation. All procedures were performed with the patient under conscious sedation.

Indications for examination were: weight loss of unclear etiology, elevated carbohydrate-associated antigen (CA) 19-9, painless jaundice, acute pancreatitis of un- known etiology, pancreatic hyperenzymemia (increase of either lipase or amylase value or both), “double duct sign” upon imaging (present in transabdominal sonography, MRI or CT), a solid mass suspected upon cross- sectional imaging or transabdominal sonography.

A solid pancreatic lesion was defined as a lesion of the pancreatic parenchyma with predominantly solid tissue seen at standard/grey scale EUS (ie, the solid component constituted at least 90% of the total volume of the lesion). The presence of cystic lesions in other parts of

Fig 1. Flow chart of the study group, rate of surgery and in- flammatory morphological changes in EUS

(3)

the pancreas was not an exclusion criterion. When an intraductal papillary mucinous neoplasia (IPMN) was not suspected preoperatively and diagnosed only through the histopathological examination, this was reported in the final data.

For the analysis, the patient collective was divided into two groups according to the presence or absence of sign for pancreatitis at EUS in the surrounding pancreatic parenchyma all around the focal lesion: Group 1:

healthy pancreas (HP) around the focal lesion and Group 2: inflamed pancreas (IP) around the focal lesion.

Patients presenting with normal pancreatic parenchyma upon EUS were assigned to the HP group. Patients presenting with signs of acute [36] or chronic pancreatitis at EUS according to the Rosemont criteria for chronic pancreatitis (i.e. enlarged hypoechoic or inhomogeneous and coarse/lobulated parenchyma, with protein plugs, fibrotic and hyperechoic spots) were assigned to the IP group [22,37-39].

Exclusion criteria were: acute biliary pancreatitis and pre-EUS diagnosis of choledocholithiasis, predominantly cystic mass, primary cystic lesions without solid component, preoperative diagnosis of IPMN.

For patients presenting with mass-forming chronic pancreatitis, cross-sectional imaging pictures were re- viewed by an expert radiologist (M.H.) specialized in abdominal imaging. Criteria of paraduodenal pancreatitis were systematically applied in order to achieve a diagnosis “a posteriori” in accordance to literature [32,33]

(thickening of the duodenal wall, fibrous tissue within the pancreaticoduodenal groove showing as hypoen- hancement of the duodenal wall after contrast medium application, signs of duodenal stenosis with/without gas- tric outlet obstruction, duodenal wall cysts, leftward dis- placement of a normal appearing gastroduodenal artery).

The radiologist was blinded to clinical data. Patients’

imaging data were obtained from the computerized ar- chive of the Department of Radiology of the University of Erlangen-Nuremberg.

The diagnostic gold standard was histology from a surgically resected specimen or FNA. Where histology could not be obtained or was not diagnostic, clinical follow-up for at least 6 months was requested in patients without a definitive diagnosis.

Statistical analysis

Parametric variables are reported as mean and standard deviation. Student’s t-test was used to compare continuous variables. The χ2 analysis was used for cat- egorical variables. The Fisher exact probability test was used for the 2 x 2 contingency tables, where suitable.

A two-tailed distribution was used with p-values <0.05 considered to be statistically significant. Statistics were

calculated using the IBM SPSS 20 statistical program (IBM Corporation, Armonk, New York, USA). Regard- ing diagnostic sensitivity and specificity of EUS predict- ing malignancy, any neoplastic lesion (e.g., pancreatic neuroendocrine tumor, distal cholangiocarcinoma, ampullary tumor) was considered as a malignancy. The diagnostic gold standard was a surgically resected specimen or diagnostic histology obtained via biopsy.

Results

Patients’ characteristics

Among all patients that underwent EUS in our institution for the suspicion of pancreatic malignancy, 218 patients were included (124 men, 94 women; mean age 60±14 years, range 20-88). The final diagnosis of a benign lesion was made in 120 patients (55.0%) and a malignant lesion was diagnosed in 98 (45%) patients (fig 2). 54 patients (24.8% of all lesions) underwent surgery and 4 patients (1.8%) refused surgery. 14 patients (25.9%) who underwent surgery because of a suspicion of malignancy (no preoperative histology) could not be confirmed in the operative specimen. EUS follow-up was performed in 61 patients (28%), follow-up 16.5±27.3 months. In 18 patients under follow-up, surgery was re- quired. Follow-up was refused by 4 patients. One patient refused both follow-up and surgery. Sensitivity of EUS for malignancy for the whole cohort was 91% and specificity was 89.2%.

EUS in patients (n=121) with a normal pancreatic parenchyma surrounding the suspicious lesion (HP-group)

Of all the 218 patients, 121 presented with a not-inflamed, normal pancreatic parenchyma in the surround-

Fig 2. Flow chart of the study group divided into malignant and benign lesions

(4)

ing tissue of the unclear focal mass (fig 3, fig 4). Clinical characteristics of the patients of the HP-group are shown in table I according to the final diagnosis of malignant and benign lesion. Thirty-three of these patients (28.9%) underwent surgery due to the suspicion of malignancy;

3 patients (2.5%) refused surgery, whereas 21 patients (17.4%) underwent EUS follow-up. The mean follow-up was 11±27.5 (1-125) months after the initial EUS examination.

EUS in the HP group showed sensitivity for malignancy of 91.4% and a specificity of 97.7%. EUS-FNA had sensitivity for malignancy of 70% and a specificity of 80%.

EUS in patients (n=97) with signs of an inflamed pancreas around the suspicious pancreatic lesion (IP-group)

Ninety-seven patients showed signs of inflammation in the pancreatic parenchyma upon EUS [69 (71.1%) men; 57.4±14.2 years]. Of these, 14 patients (14.4%) had pancreatic cancer, 36 (37.1%) chronic pancreatitis, 3 (3.1%) AIP (fig 5), 20 (20.6%) acute pancreatitis, 7 (7.2

%) necrosis, 7 paraduodenal pancreatitis (7.2 %), 3 neuroendocrine tumors (NET) (3.1%) and 4 cholangiocellular carcinoma (CCC) (4.1%). In two patients, no focal lesion was detectable on EUS (one patient showed an atrophic pancreatic parenchyma, the other one had unspecific chronic signs of pancreatitis). The clinical characteristics of patients in the IP-group comparing the subgroups of malignant and benign lesions can be seen in Table II.

In the IP-group, 40 patients (41.2%) underwent EUS follow-up [mean follow-up 19±27.2 (1-123) months after the initial EUS examination].

Fig 3. Endosonography in a case with pancreatic cancer. PD pancreatic duct, PCA pancreatic cancer, PV portal vein, CBD common bile duct

Fig 4. Endosonography in a case with a 21 mm pancreatic mass (arrow) – in histology neuroendocrine tumor. No pancreatic inflammation

Table I. Clinical characteristics of patients with normal pancreatic parenchyma surrounding the suspicious lesion (HP group)

HP group Benign

n = 44 Malignant

n = 77 p

Sex M 16 39 ns

Mean Age (years) 60.5±14.4 63±13.5 ns

BMI (kg/m²) 26.6±5.2 25.5±4.1 ns

Nicotine abuse 9 17 ns

Alcohol abuse 13 23 ns

Exocrine failure 3 9 ns

CA 19-9 elevated 2 34 0.001

IgG4 elevated 5 0 0.016

Mass on EUS 18 70 <0.001

Mass on CT 18 57 0.001

Synchronous Neoplasm 3 6 ns

Extrapancreatic

involvement 2 22 0.001

Enlarged LN 6 30 0.003

No lesion on EUS 25 6 <0.001

Double duct sign 2 20 0.003

MPD dilation 4 37 <0.001

MPD Stenosis 5 18 ns

Bile duct dilation 6 28 0.004

Lipomatosis 15 9 0.003

Jaundice 4 20 0.025

Hyperlipasemia 4 13 ns

Weight loss 8 30 0.016

Diabetes 5 22 0.04

Surgery 3 32 <0.001

The results are expressed as number or mean±standard deviation.

BMI – body mass index; MPD – main pancreatic duct; ns – not significant; FNA fine needle aspiration; TUS – transabdominal ultrasonography; LN – lymph nodes; CA 19-9: carbohydrate-associated antigen 19-9. Exocrine Failure was defined as pancreatic enzyme supplementation therapy and/or decreased values of pancreatic elastase on stool spot (ELISA assay)

(5)

EUS in the IP-group showed sensitivity for malignancy of 44% and a specificity of 87.1%. EUS-FNA presented sensitivity for malignancy of 40% and a specificity of 20%.

In Table III the different etiologies of the pancreatic lesions are presented after proof of the final diagnosis and follow-up, again divided according to HP and IP.

Comparison between IP-group and HP-group Patients in the IP-group showed different clinical features compared to patients in the HP-group (Table IV).

Patients in the IP group were more often male with a medical history of nicotine and alcohol abuse, presented more often pancreatic calcification and a pancreatic mass that could not be reproduced in EUS or disappeared during follow-up, indicating the inflammatory genesis of the Fig 5. Endosonography in a case with focal autoimmune pan-

creatitis. PV Portal Vein, *pancreatic focal mass, CBD common bile duct. Signs of inflamed pancreatic parenchyma

Table II. Clinical characteristics of patients with signs of pancreatic inflammation (IP group) in the pancreatic parenchyma surrounding the suspicious lesion

IP group Benign

n = 76 Malignant

n = 21 p

Sex M 56 13 ns

Age (years) 56.2±14.1 61.6±14.3 ns

BMI (kg/m²) 25.0±4.5 25.6±4.9 ns

Nicotine abuse 43 9 ns

Alcohol abuse 46 11 ns

CA 19-9 elevation 9 10 0.008

IgG4 elevation 0 1 ns

Mass 18 70 <0.001

Extrapancreatic

involvement 6 5 0.056

Synchrone Neoplasm 5 3 ns

LN 35 11 ns

No lesions in EUS 41 3 0.001

Double duct sign 7 8 0.001

MPD dilation 31 15 0.013

MPD Stenosis 16 8 ns

Bile duct dilation 11 9 0.03

Lipomatosis 5 3 ns

Jaundice 9 9 0.001

Hyperlipasemia 31 6 ns

Weight loss 22 10 ns

Acute pancreatitis 31 2 0.007

Diabetes 26 5 ns

Surgery 11 8 0.05

Calcifications 39 10 ns

M – male; BMI – body mass index; MPD – main pancreatic duct;

ns – not significant; FNA fine needle aspiration; TUS – transabdominal ultrasonography; LN – lymph nodes; CA 19-9: carbohydrate-associated antigen 19-9. Exocrine Failure was defined as pancreatic enzyme supplementation therapy and/or decreased values of pancreatic elastase on stool spot (ELISA assay).

Table III. Etiological distribution of focal pancreatic lesions in consideration of presence and absence of echo-morphological signs of pancreatic inflammation

Aetiology Healthy pancreas (HP)n = 121 (55.5%)

Inflamed pancreas (IP)n = 97 (44.5 %)

p

Pancreatic cancer 46 (21.1) 14 (6%) 0.038

NET 23 (10.5) 3 (0.1%) <0.001

Ampullary

carcinoma 4 (0.1) 0 (0%) ns

CCC 3 (0.1) 4 (0.1%) ns

IPMN 1 (0.5) 1 (0.5%) ns

Acute

pancreatitis 0 20 (9%) <0.001

Chronic

pancreatitis 0 36 (16.5%) <0.001

Pancreatic

necrosis 0 7 (3.2%) <0.001

Autoimmune

pancreatitis 8 (3.7) 3 (1.3%) ns

Paraduodenal

pancreatitis 1 (0.5) 7 (3.2%) 0.01

Choledocholithiasis 1(0.5) 0 ns

Ventral/dorsal

split 6 (2.7) 0 <0.001

Pancreatic

lipomatosis 3 (1) 0 ns

Enlarged

lymphnodes 2 (1) 0 ns

Microcystic serous mucinous neoplasm

1 (0.5) 0 ns

Other* 22 (10) 2 (1) <0.001

Results are expressed as number (%). NET – neuroendocrine tumors; CCC – cholangiocellular carcinoma; IPMN – intraductal papillary mucinous neoplasia. *Other rare conditions such as benign choledochus strictures, papillitis stenosans and patients with a pancreatic occupying lesion who refused follow-up

(6)

lesion. In addition, they presented more often acute pancreatitis and hyperlipasemia.

Patients with AIP and paraduodenal pancreatitis were found in both groups. Clinical and imaging features of

these patients are shown in supplementary Table I and II, on the journal site. In our case IgG4 elevation presented a sensitivity of 83.3% and a specificity of 77.8% for AIP.

Discussions

Our findings describe the influencing factors in the management of solid pancreatic masses undergoing EUS. The accuracy of EUS, with or without FNA, for diagnosing pancreatic cancer in specific cohorts has been described earlier [40,41]. Our data represent a “real-life”

scenario from a tertiary referral center in contrast. Ad- ditionally, we focused on the characteristics of the surrounding pancreatic parenchyma beside the solid mass with a discrimination of an inflamed or normal pancreatic tissue. In our study, the sensitivity of EUS regarding the diagnosis of malignancy in IP-group was 44% and signif- icantly lower than the 91.4% observed in the HP-group.

Sensitivities between 54% and 74% have been reported in previous studies evaluating solid pancreatic masses with EUS-FNA in the setting of chronic pancreatitis [42-44].

Moreover, pancreatic calcifications can create acoustic shadows that may prevent a complete evaluation of the whole gland. Also, dilation of the main pancreatic duct (MPD) may occur years before the detection of a small neoplasm in patients with normal pancreatic parenchyma [45], whereas in the setting of chronic pancreatitis, MPD fibrotic dilation may be present without a cause of obstruction [46].

The differentiation between pancreatic cancer and a solid lesion due to mass forming pancreatitis is still challenging. According to our results, also in the setting of normal pancreatic parenchyma, AIP and paraduodenal pancreatitis can be the causes of focal lesions. Ventral/

dorsal split leading to suspicion of cancer, which may appear as a small hypoechoic area in the posterior part of the pancreatic head, represents a differential diagnosis, which can be excluded by experienced endosonographers [47]. In unclear cases, repeated examinations in intervals of at most 4-12 weeks are recommended [48].

Pancreatic lipomatosis, especially if distributed unevenly through the gland, can be confused with pancreatic cancer in cross-sectional imaging [49]. In our series, EUS represents an excellent tool for diagnosis, since no patient with focal lipomatosis underwent pancreatic surgery for suspected malignancy.

The number of FNA procedures in the pancreas is much lower in our center compared to other studies cited in the published literature [50]. The explanation may be that there is a tendency to perform resection in the presence of a pancreatic mass rather than preoperative diagnosis, since the probability of malignancy is high (more Table IV. Clinical comparison between the patients with nor-

mal pancreatic parenchyma surrounding the suspicious lesion (HP group) and patients with signs of pancreatic inflammation (IP group)

Patients

(n = 218) HP-group

(n = 121) IP-group (n = 97) p

Sex (M=124) 55 69 P<0.01

Age (mean, years) 62±13.7 57.4±14.2 0.016

BMI (kg/m²) 25.9 25.2 ns

Nicotine abuse 26 52 P<0.01

Alcohol abuse 36 57 P<0.01

Mass 88 51 0.002

Extrapancreatic

involvement 24 11 ns

LN 36 46 0.01

No Lesions in EUS 31 44 0.02

Dilation MPD 41 46 0.048

Stenosis MPD 23 24 ns

Dilation common Bile

duct 34 20 ns

Double duct sign 22 15 ns

Follow-up (FU) 21 40 <0.001

Months FU 11±27.5 19±27.2 ns

Histology obtained

preoperatively 44 12 <0.001

Surgery 35 19 ns

Surgery refused 3 1 ns

Synchronous neoplasm 9 8 ns

Diabetes 27 31 ns

Acute pancreatitis 8 33 <0.001

Weight loss 38 32 ns

Hyperlipasemia 17 37 <0.001

Jaundice 24 18 ns

Pancreatic

calcifications 0 49 <0.001

Pancreatic lipomatosis 24 8 0.016

Chronic pancreatitis 0 79 <0.001

CA 19-9 elevation 36 19 ns

IgG4 elevation 5 1 0.025

Mass in CT 75 45 ns

Malignancy 76 21 <0.01

BMI – body mass index; MPD – main pancreatic duct; ns – not significant; FNA fine needle aspiration; TUS – transabdominal ultrasonography; LN – lymph nodes; CA 19-9: carbohydrate-associated antigen 19-9. Exocrine Failure was defined as pancreatic enzyme supplementation therapy and/or decreased values of pancreatic elastase on stool spot (ELISA assay).

(7)

than 90%) [18]. Furthermore, a negative FNA would not necessarily avoid resection due to the possibility of false negative findings. Preoperative tissue acquisition may be considered when an inflammatory mass is suspected on the basis of clinical history (familiarity for pancreatitis, young age, alcohol and smoking habits, previous diagnosis of autoimmune diseases) and of radiological findings suggestive for inflammation and not for cancer. On the other hand, recently a surgical multicentric series from the Netherlands reported results for EUS–FNA similar to ours; i.e. they reported a sensitivity of 63% and specificity of 61.5% for detecting pancreatic malignancy [51] in a “real-life” cohort.

FNA in our series did not assist us in diagnosing benign lesions such as AIP and paraduodenal pancreatitis, since alternative means for tissue diagnosis had to be em- ployed. In unclear cases, clinical and imaging follow-up with more than one modality (usually EUS with CT) or transabdominal core biopsies were necessary for achiev- ing a definitive diagnosis. After exclusion of malignancy with invasive means, as suggested for ICDC in cases suspicious of AIP, a steroid trial of 2 weeks with repeated imaging (MRI) in the short interval may be very helpful for a final diagnosis, despite ambiguous histology findings. In cases of paraduodenal pancreatitis, the history of intense alcohol consumption and smoking may raise suspicions of the disease and define an appropriate treat- ment, which may also lead to a pancreatoduodenectomy [34]. However, in the case of cancer the operation should be performed within 1 month, whereas in the case of paraduodenal pancreatitis surgery can be delayed.

According to our results, there were clinical differ- ences between the groups. Among the IP-group, patients were more often males, smokers and drinkers. They also presented more often enlarged lymph nodes and diabetes.

Among the IP group the pancreatic mass occasionally re- solved during follow-up which indicates that pancreatic inflammations (showed also from the increase of lipase in the serum) play a confusing role in the diagnostic process with EUS.

The presence of a pancreatic mass on EUS is not in all cases an indicator of cancer. The presence of enlarged lymph nodes in the absence of pancreatitis is suggestive for locally disseminated cancer. In contrast to the IP- group, a dilation of the pancreatic duct, bile duct or both (double duct sign) and the clinical presence of jaundice, weight loss and worsening diabetes are alarm signs and should lead to aggressive tissue sampling and/or operative evaluation. As already reported by Alston et al we also confirm that particularly in patients of the HP-group, the presence of a mass and weight loss were more often associated with malignancy [52].

This study has some limitations. First, the patients included were a heterogeneous group due to the retrospective design of the study and no effort could be made to quantify the degree of “suspicion” for the presence of pancreatic cancer. However, the patient collective represents a “real-life” cohort of our clinical practice, without selections for high prevalence of malignant neoplasms.

As a typical example, the study by Mallery et al had a 92% prevalence of malignancy [53]. Secondly, the patient collective was gathered from our EUS prospective database, so there may be a selection bias for lesions with untypical imaging appearance that requires further investigation before surgery. In our institution, patients are usually discussed in an interdisciplinary conference and are directly referred to surgery if the suspicion of malignancy is strong. Thirdly, not all patients underwent surgical resection, so a radical surgical/pathological diagnosis could not be achieved for all patients.

Conclusions

Sensitivity and specificity of EUS for malignancy is lower in the presence of pancreatitis. In the setting of an inflamed pancreas, an obstructive etiology causing pancreatitis should be expected and clinically further investigated with patient history, imaging and laboratory data (i.e. cancer or chronic inflammation due to smoking, alcohol abuse or autoimmunity). If the pancreas is sono- morphologically normal, a careful evaluation to recog- nize small pancreatic and extrapancreatic pathology not involving the pancreatic duct (such as anatomic variant or small pancreatic neoplasms) should be performed.

Also, in the setting of a normal pancreatic gland, focal mass forming pancreatitis as AIP and paraduodenal pancreatitis should be considered. Unclear cases should thus undergo early follow-up examinations by EUS with ad- ditional complementing imaging modalities.

Conflict of interest: none Acknowledgment:

The present work was performed in fulfillment of the requirements for obtaining the degree ‘Dr. med.’ at the Friedrich-Alexander-University Erlangen-Nurnberg (FAU), Germany.

References

1. DeWitt J, Devereaux B, Chriswell M, et al. Comparison of endoscopic ultrasonography and multidetector computed tomography for detecting and staging pancreatic cancer.

Ann Intern Med 2004;141:753-763.

(8)

2. Ardengh JC, de Paulo GA, Ferrari AP. Pancreatic carcino- mas smaller than 3.0 cm: endosonography (EUS) in diagnosis, staging and prediction of resectability. HPB (Oxford) 2003;5:226-230.

3. Dietrich CF, Sahai AV, D’Onofrio M, et al. Differential diagnosis of small solid pancreatic lesions. Gastrointest En- dosc 2016;84:933-940.

4. Agarwal B, Abu-Hamda E, Molke KL, Correa AM, Ho L.

Endoscopic ultrasound-guided fine needle aspiration and multidetector spiral CT in the diagnosis of pancreatic cancer. Am J Gastroenterol 2004;99:844-850.

5. Yarandi SS, Runge T, Wang L, et al. Increased Incidence of Benign Pancreatic Pathology following Pancreaticoduo- denectomy for Presumed Malignancy over 10 Years despite Increased Use of Endoscopic Ultrasound. Diagn Ther En- dosc 2014;2014:701535.

6. de la Fuente SG, Ceppa EP, Reddy SK, Clary BM, Tyler DS, Pappas TN. Incidence of benign disease in patients that underwent resection for presumed pancreatic cancer diagnosed by endoscopic ultrasonography (EUS) and fine- needle aspiration (FNA). J Gastrointest Surg 2010;14:1139- 1142.

7. Sasson AR, Gulizia JM, Galva A, Anderson J, Thompson J.

Pancreaticoduodenectomy for suspected malignancy: have advancements in radiographic imaging improved results?

Am J Surg 2006;192:888-893.

8. Abraham SC, Wilentz RE, Yeo CJ, et al. Pancreaticoduo- denectomy (Whipple resections) in patients without malignancy: are they all ‘chronic pancreatitis’? Am J Surg Pathol 2003;27:110-120.

9. Adsay NV, Basturk O, Klimstra DS, Klöppel G. Pancreatic pseudotumors: non-neoplastic solid lesions of the pancreas that clinically mimic pancreas cancer. Semin Diagn Pathol 2004;21:260-267.

10. Barens SA, Lillemoe KD, Kaufman HS, et al. Pan- creaticoduodenectomy for benign disease. Am J Surg 1996;171:131-134.

11. Cohen JR, Kuchta N, Geller N, Shires GT, Dineen P.

Pancreaticoduodenectomy for benign disease. Ann Surg 1983;197:68-71.

12. Fuente SG, Ceppa EP, Reddy SK, Clary BM, Tyler DS, Pap- pas TN. Incidence of benign disease in patients that underwent resection for presumed pancreatic cancer diagnosed by endoscopic ultrasonography (EUS) and fine-needle aspiration (FNA). J Gastrointest Surg 2010;14:1139-1142.

13. Gomes RM, Bal M, Patkar S, Goel M, Shrikhande SV. Un- expected benign histopathology after pancreatoduodenectomy for presumed malignancy: accepting the inevitable.

Langenbecks Arch Surg 2016;401:169-179.

14. Kavanagh DO, O’Riain C, Ridgway PF, et al. Radical pancreaticoduodenectomy for benign disease. ScientificWorld- Journal 2008;8:1156-1167.

15. Smith CD, Behrns KE, van Heerden JA, Sarr MG. Radical pancreatoduodenectomy for misdiagnosed pancreatic mass.

Br J Surg 1994;81:585-589.

16. van Gulik TM, Reeders JW, Bosma A, et al. Incidence and clinical findings of benign, inflammatory disease in patients

resected for presumed pancreatic head cancer. Gastrointest Endosc 1997;46:417-423.

17. van Heerde MJ, Biermann K, Zondervan PE, et al. Preva- lence of autoimmune pancreatitis and other benign disor- ders in pancreatoduodenectomy for presumed malignancy of the pancreatic head. Dig Dis Sci 2012;57:2458-2465.

18. Vitali F, Hansen T, Kiesslich R, et al. Frequency and characterization of benign lesions in patients undergoing surgery for the suspicion of solid pancreatic neoplasm. Pan- creas 2014;43:1329-1333.

19. Takahashi H, Moslim MA, Presser N, et al. Absence of a Periampullary Mass on Cross-sectional Imaging Delays Di- agnosis and Time to Pancreatoduodenectomy But Does Not Impair Outcome. J Gastrointest Surg 2016;20:1179-1187.

20. Kamata K, Kitano M, Omoto S, et al. New endoscopic ultrasonography techniques for pancreaticobiliary diseases.

Ultrasonography 2016;35:169-179.

21. Lee YN, Moon JH, Kim HK, et al. Core biopsy needle ver- sus standard aspiration needle for endoscopic ultrasound- guided sampling of solid pancreatic masses: a randomized parallel-group study. Endoscopy 2014;46:1056-1062.

22. Piscaglia F, Nolsoe C, Dietrich CF, et al. The EFSUMB Guidelines and Recommendations on the Clinical Practice of Contrast Enhanced Ultrasound (CEUS): update 2011 on non-hepatic applications. Ultraschall Med 2012;33:33-59.

23. Saftoiu A, Dietrich CF, Vilmann P. Contrast-enhanced harmonic endoscopic ultrasound. Endoscopy 2012;44:612- 24. Dietrich CF, Saftoiu A, Jenssen C. Real time elastography 617.

endoscopic ultrasound (RTE-EUS), a comprehensive review. Eur J Radiol 2014;83:405-414.

25. Kitano M, Kudo M, Yamao K, et al. Characterization of small solid tumors in the pancreas: the value of contrast- enhanced harmonic endoscopic ultrasonography. Am J Gastroenterol 2012;107:303-310.

26. Seicean A, Badea R, Moldovan-Pop A, et al. Harmonic Contrast-Enhanced Endoscopic Ultrasonography for the Guidance of Fine-Needle Aspiration in Solid Pancreatic Masses. Ultraschall Med 2017;38:174-182.

27. Cieslak KP, van Santvoort HC, Vleggaar FP, et al. The role of routine preoperative EUS when performed after contrast enhanced CT in the diagnostic work-up in patients suspected of pancreatic or periampullary cancer. Pancreatology 2014;14:125-130.

28. Asbun HJ, Conlon K, Fernandez-Cruz L, et al. When to perform a pancreatoduodenectomy in the absence of posi- tive histology? A consensus statement by the International Study Group of Pancreatic Surgery. Surgery 2014;155:887- 29. Krautz C, Nimptsch U, Weber GF, Mansky T, Grützmann 892.

R. Effect of Hospital Volume on In-hospital Morbidity and Mortality Following Pancreatic Surgery in Germany. Ann Surg 2018;267:411-417.

30. Shimosegawa T, Chari ST, Frulloni L, et al. International consensus diagnostic criteria for autoimmune pancreatitis:

guidelines of the International Association of Pancreatol- ogy. Pancreas 2011;40:352-358.

(9)

31. Adsay NV, Zamboni G. Paraduodenal pancreatitis: a clini- co-pathologically distinct entity unifying “cystic dystrophy of heterotopic pancreas”, “para-duodenal wall cyst”, and

“groove pancreatitis”. Semin Diagn Pathol 2004;21:247- 32. Vullierme MP, Vilgrain V, Flejou JF, et al. Cystic dystro-254.

phy of the duodenal wall in the heterotopic pancreas: ra- diopathological correlations. J Comput Assist Tomogr 2000;24:635-643.

33. Procacci C, Graziani R, Zamboni G, et al. Cystic dystrophy of the duodenal wall: radiologic findings. Radiology 1997;205:741-747.

34. Arvanitakis M, Rigaux J, Toussaint E, et al. Endotherapy for paraduodenal pancreatitis: a large retrospective case series. Endoscopy 2014;46:580-587.

35. de Pretis N, Capuano F, Amodio A, et al. Clinical and Mor- phological Features of Paraduodenal Pancreatitis: An Italian Experience With 120 Patients. Pancreas 2017;46:489-495.

36. Jenssen C, von Lampe B, Kahl S. Endoscopic Ultrasound in Acute Pancreatitis. Video Journal and Encyclopedia of GI Endoscopy 2013;1:554–559.

37. Catalano MF, Sahai A, Levy M, et al. EUS-based criteria for the diagnosis of chronic pancreatitis: the Rosemont classification. Gastrointest Endosc 2009;69:1251-1261.

38. Trikudanathan G, Munigala S, Barlass U, et al. Evaluation of Rosemont criteria for non-calcific chronic pancreatitis (NCCP) based on histopathology - A retrospective study.

Pancreatology 2017;17:63-69.

39. Chronic Pancreatitis German Society of Digestive and Metabolic Diseases (DGVS), Hoffmeister A,Mayerle J, et al. S3-Consensus guidelines on definition, etiology, diagnosis and medical, endoscopic and surgical management of chronic pancreatitis German Society of Digestive and Met- abolic Diseases (DGVS). Z Gastroenterol 2012;50:1176- 1224.

40. Wang W, Shpaner A, Krishna SG, et al. Use of EUS-FNA in diagnosing pancreatic neoplasm without a definitive mass on CT. Gastrointest Endosc 2013;78:73-80.

41. Hewitt MJ, McPhail MJ, Possamai L, Dhar A, Vlavianos P, Monahan KJ. EUS-guided FNA for diagnosis of solid pancreatic neoplasms: a meta-analysis. Gastrointest Endosc 2012;75:319-331.

42. Bang JY, Varadarajulu S. Neoplasia in chronic pancreatitis:

how to maximize the yield of endoscopic ultrasound-guided fine needle aspiration. Clin Endosc 2014;47:420-424.

43. Fritscher-Ravens A, Brand L, Knofel WT, et al. Compari- son of endoscopic ultrasound-guided fine needle aspira-

tion for focal pancreatic lesions in patients with normal parenchyma and chronic pancreatitis. Am J Gastroenterol 2002;97:2768-2775.

44. Varadarajulu S, Tamhane A, Eloubeidi MA. Yield of EUS-guided FNA of pancreatic masses in the presence or the absence of chronic pancreatitis. Gastrointest Endosc 2005;62:728-736.

45. Gangi S, Fletcher JG, Nathan MA, et al. Time interval between abnormalities seen on CT and the clinical diagnosis of pancreatic cancer: retrospective review of CT scans obtained before diagnosis. AJR Am J Roentgenol 2004;182:897-903.

46. Krishna NB, Mehra M, Reddy AV, Agarwal B. EUS/EUS- FNA for suspected pancreatic cancer: influence of chronic pancreatitis and clinical presentation with or without obstructive jaundice on performance characteristics. Gastro- intest Endosc 2009;70:70-79.

47. Savides TJ, Gress FG, Zaidi SA, Ikenberry SO, Hawes RH. Detection of embryologic ventral pancreatic parenchyma with endoscopic ultrasound. Gastrointest Endosc 1996;43:14-19.

48. Bhutani MS, Gress FG, Giovannini M, et al. The No En- dosonographic Detection of Tumor (NEST) Study: a case series of pancreatic cancers missed on endoscopic ultrasonography. Endoscopy 2004;36:385-389.

49. Gerritsen A, Bollen TL, Nio CY, et al. Diagnostic value of a pancreatic mass on computed tomography in patients undergoing pancreatoduodenectomy for presumed pancreatic cancer. Surgery 2015;158:173-182.

50. Eloubeidi MA, Varadarajulu S, Desai S, et al. A prospective evaluation of an algorithm incorporating routine preoperative endoscopic ultrasound-guided fine needle aspiration in suspected pancreatic cancer. J Gastrointest Surg 2007;11:813-819.

51. Gerritsen A, Molenaar IQ, Bollen TL, et al. Preoperative characteristics of patients with presumed pancreatic cancer but ultimately benign disease: a multicenter series of 344 pancreatoduodenectomies. Ann Surg Oncol 2014;21:3999- 4006.

52. Alston EA, Bae S, Eltoum IA. Suspicious cytologic diagnostic category in endoscopic ultrasound-guided FNA of the pancreas: Follow-up and outcomes. Cancer Cytopathol 2016;124:53-57.

53. Mallery JS, Centeno BA, Hahn PF, Chang Y, Warshaw AL, Brugge WR. Pancreatic tissue sampling guided by EUS, CT/US, and surgery: a comparison of sensitivity and specificity. Gastrointest Endosc 2002;56:218-224.