The study is based on a multi-center approach of needle based confocal laser endomicroscopy (nCLE) combined with endoscopic ultrasound (EUS) and EUS-guided fine needle aspiration (FNA) to evaluate pancreatic cystic lesions (PCL), in order to obtain a correct histopathological diagnosis.After detection of PCL, certain morphological EUS features allow the discrimination of specific cyst types. Additionally, EUS-FNA is recommended as the first-line procedure whenever pathological diagnosis is required; however the procedure has its drawbacks, mainly represented by the relatively low negative predictive value in diagnosing pancreatic cancer. In this case a more precisely diagnostic tool is required; the potential role of CLE has been explored in gastrointestinal (GI) pathology showing good accuracy for predicting the final histopathological diagnosis based on immediate evaluation of tissue and vascular patterns. Although the clinical impact of nCLE for the decision making algorithms in cystic pancreatic neoplasm has not yet been described, the hypothesis is that EUS-nCLE could allow targeted tissue sampling of cystic pancreatic neoplasms resulting in more accurate diagnosis. The aim of the study is to describe the clinical impact of nCLE for the clinical decision management algorithm based on EUS, EUS-FNA and/or EUS-CLE imaging criteria for cystic pancreatic neoplasms, while evaluating also the feasibility and safety of nCLE examination.

Clinical Impact of Needle-based Confocal Laser Endomicroscopy of Cystic Pancreatic Lesions (CINE-Cyst)

Background The diagnostic evaluation of cystic pancreatic lesions represents a difficult problem for clinical decision making, due to the increased incidental discovery of such lesions during cross-sectional imaging (computed tomography or magnetic resonance). Prevalence of cystic pancreatic lesions is estimated between 2 to 20% in the adult population (1-3), while autopsy studies show that the prevalence increases with age (4). The WHO classification (revised in 2010) reported several categories, with serous cystadenoma (SCA), mucinous cystic neoplasms (MCN), intraductal papillary mucinous neoplasms (IPMN) and pseudopapillary neoplasms representing more than 90% (5). Several of these lesions progress to cystadenocarcinoma (with the notable exception of SCA), hence the differential diagnosis is extremely important for the clinical decisions that currently involve either imaging follow-up (usually based on EUS and magnetic resonance imaging [MRI]) or referral to surgery with a final pathological diagnosis established after resection.

Nevertheless, there is considerable overlap in the imaging characteristics of benign, premalignant or malignant pancreatic cysts which leads to unclear clinical management algorithms. Several international guidelines addressed the diagnosis and management of cystic pancreatic lesions (6-8). The most comprehensive guideline reported a clinical decision algorithm which included clinical and imaging criteria, like pancreatitis / jaundice or cyst size > 3 cm, thickening / enhancing cyst walls, main pancreatic duct (MPD) size 5-9 mm, non-enhancing mural nodules and abrupt changes in the caliber of pancreatic duct with distal pancreatic atrophy [8]. If any of these clinical / imaging characteristics are present then EUS is considered mandatory to assess a definite mural nodule, main duct features suspicious for involvement, as well as to perform EUS-guided fine needle aspiration (FNA) with cytopathology exam in order to exclude malignancy.

Endoscopic ultrasound (EUS) represents a highly valuable tool in the management of patients with focal pancreatic masses, either solid or cystic. As a minimally invasive technique that enables high-resolution imaging of the pancreatic parenchyma and surrounding structures it is considered the method of choice for the detection of clinically suspected pancreatic lesions, with a negative predictive value close to 100% [9]. Its diagnostic sensitivity was shown by previous studies to be superior compared to other imaging techniques, especially in the case of smaller tumors [10, 11]. After detection of cystic pancreatic lesions, certain morphological features allow the discrimination of specific types [12]:

• pseudocysts - parenchymal changes , without septation or nodules
• serous cystadenomas - multiple microcysts and honeycomb appearance
• mucinous cystadenoma - thickened septations and wall, with calcifications
• intrapapillary mucinous neoplasia (IPMN) - communication with the MPD There aren't universally accepted morphological parameters to predict either the type or the risk of malignancy by EUS [8]. "Worrisome features" on imaging include cyst size (above 3 cm), thickened cyst walls, MPD size of 5-9 mm, non-enhanced mural nodules, abrupt changes in MPD caliber with distal pancreatic atrophy, and lymphadenopathy. "High risk stigmata" include obstructive jaundice in a cystic pancreatic head mass, enhanced solid component or MPD size of over 10 mm, thus warranting an indication for surgery, if clinically feasible. Generally, all cysts with "worrisome features" and cysts over 3 cm, without worrisome features, should undergo EUS, while all cysts with "high-risk stigmata" should undergo surgical resection [8].

Additionally, EUS enables guided fine needle aspiration (EUS-FNA) which is currently recommended as the first-line procedure whenever pathological diagnosis is required [13]. However, EUS-FNA as a sampling technique has its drawbacks, mainly represented by the relatively low negative predictive value in diagnosing pancreatic cancer. It thus cannot reliably rule out a diagnosis of malignancy and patients with high clinical suspicion usually need repeated FNA [14]. The role of EUS-FNA for the differential diagnosis of focal cystic lesions is still controversial, as CEA and amylase levels as well as cytology are often useful to differentiate various lesions, but not all of them [8]. Thus, SCA have low levels of both CEA and amylase. For CEA, values over 192-200 ng/ml have 80% diagnostic accuracy for mucinous cysts [15]. Likewise, amylase levels might be increased in both IPMN and MCN, making the differential diagnosis difficult [16]. Cytology has a limited sensitivity, as the aspirates yield a low number of cells, being most useful if it detects atypical epithelial cells which can highly predict malignancy, again with an 80% accuracy [17].

Confocal laser endomicroscopy (CLE) has emerged in recent years as a novel technique that actually enables in vivo microscopic analysis during ongoing endoscopy. Endomicroscopy can be performed either with dedicated (eCLE) or with miniprobe-based systems (pCLE) [18]. It is a contrast based technique, the most widely used agent being the intravenously administered fluorescein [19]. The probe-based endomicroscopy system consists of a flexible catheter probe representing a bundle of optical fibers linked to a micro-objective, a laser scanning unit and the control and acquisition software (Cellvizio; Mauna Kea Technology, Paris, France). The flexible confocal miniprobes were specifically designed to be passed through the working channels of standard endoscopes, biliary catheters or cholangioscopes and thus the pCLE system can be easily integrated in any endoscopy unit. The principle of the technique is based on a laser beam of defined wavelength being focused towards the targeted tissue and the recaptured signal is displayed as 'optical biopsies' in a single horizontal plane.

The potential role of CLE has been explored in pathology of both upper and lower gastrointestinal tract, showing good accuracy for predicting the final histopathological diagnosis based on immediate evaluation of tissue and vascular patterns [20]. Recently CLE has gone beyond the superficial luminal indications with the development of a new microprobe, i.e. a flexible probe thin enough that it can be passed through a 19-gauge needle. Thus under EUS guidance solid organs can be accessed for real-time microscopic information. nCLE imaging of abdominal organs has been so far achieved in animal models [21, 22]. The feasibility of the technique was also proved in a clinical study [23] and descriptive criteria for the diagnosis of pancreatic cystic neoplasms were developed from a multicentre trial [24]. However, the clinical impact of nCLE for the decision making algorithms in cystic pancreatic neoplasm has not yet been described. The hypothesis is that EUS-nCLE could allow targeted tissue sampling of cystic pancreatic neoplasms resulting in more accurate diagnosis. With further validation of the technique real-time pathological diagnosis could be obtained with immediate initiation of an adequate therapy after a single investigation.

Aim The aim of the proposed study is to describe the clinical impact of nCLE for the clinical decision management algorithm based on EUS, EUS-FNA and/or EUS-CLE imaging criteria for cystic pancreatic neoplasms, while evaluating also the feasibility and safety of nCLE examination.

Imaging tests All patients with a suspicion (clinical, US, CT/MRI) of cystic pancreatic lesions will be evaluated by EUS, EUS-FNA and EUS-nCLE and compared with the final pathological diagnosis.

• For EUS examination linear instruments will be used to perform complete examination of the pancreas.

  • Lesion characteristics (echogenicity, echostructure, size, wall, mural nodules, septations, etc.) will be described.
  • The presence of regional lymph nodes will be reported with their maximal size, echogenicity, shape and margins.
  • Identification of liver metastasis will also be looked upon.

• EUS-nCLE will be performed after EUS identification of the cystic pancreatic lesion / lymph node / liver metastasis:

  • The confocal microprobe will be preloaded in a 19G FNA needle as previously described [13] and advanced into the lesion under EUS guidance.
  • nCLE examination will follow after the intravenous administration of the contrast agent (2.5 ml fluorescein 10%).
  • Image data will be stored digitally for offline analysis (at least 2 suggestive images and 2 suggestive movies of 10 seconds, each).

• EUS-FNA will be performed after image acquisition for cytology smears and cell blocks to enable a final pathological diagnosis. CEA, CA19-9 and amylase levels will be measured in the aspirate. Molecular analyses like KRAS and GNAS will be performed, if available.

• nCLE images will be analyzed during the examination by the principal investigator, with clinical and other procedural information in mind. In a second step offline analysis, the correlations between representative CLE images and classical hematoxylin and eosin sections will be identified.

Inclusion criteria:

• Age > 18 years old, male or female
• Patients diagnosed with cystic pancreatic lesions with an indication for EUS-FNA
• Signed informed consent for EUS, EUS-FNA and EUS-nCLE performed during a single examination under sedation.

Exclusion criteria:

• Failure to provide informed consent
• Patients with a contraindication for EUS-FNA
• Known allergy to fluorescein
• Pregnant or breast-feeding patients

Data collected for each participant will include:

• Personal data (name, surname, age, sex)
• EUS variables (tumor characteristics)
• EUS-FNA results, including CEA, CA19-9, amylase levels
• EUS-nCLE images (suggestive images) and movies, digitally recorded and de-identified
• Histological and immunohistochemical findings (final diagnosis)

The study will consist of two major arms:

1. Retrospective analysis of previous prospectively included patients with cystic pancreatic neoplasms examined by EUS, EUS-FNA and EUS-nCLE with a final confirmation of the diagnosis after EUS-FNA (cytopathology) or surgery (histopathology).
2. Prospective analysis will include patients referred for EUS, EUS-FNA and EUS-nCLE of suspected cystic pancreatic lesions during a 24-months period. The indication for this investigation will be based on the patient's clinical history and previous imaging studies (abdominal ultrasound, CT scan, MRI with MRCP, etc.).