Aims and scope of the registry study

The registry was developed to provide a framework for studying the natural history of pancreatic cystic lesions. The study idea was initially established at the German Pancreas Club meeting 2020 and is endorsed and sponsored by the German Pancreas Club (DPC). Regarding the above-mentioned aim, the following aspects were especially considered by the founding team (KCH, JM, CWM, SK, RJ, PV, MD, TK, PV, SR and UFW). Relevant outcomes and patient variables were chosen by the founding team. Definitions of the International Study Group of Pancreatic Surgery (ISGPS), the TNM classification system, the German S3 guideline for the treatment of pancreatic cancer and the Fukuoka guidelines were utilized. The updated Kyoto guidelines were added August 13^th, 2024. Ethical approval was obtained from the University of Luebeck (2020-20-225)/(09/07/2020), first patient entered:14/01/2022. Participating centers have obtained approval from their respective ethics commission prior to data collection: University Medical Center Rostock (18/10/2021, first patient entered: 16/02/2023, University Medical Center Goettingen (25/10/2021, first patient entered: 24/01/2023), Technical University of Munich (26/10/2021, first patient entered: 11/07/2023), University Medical Center Halle (31/01/2022, first patient entered 23/04/2025), Ulm University Hospital (29/08/2022, first patient entered: 22/04/2024), University Medical Center Heidelberg (26/01/2024, first patient entered: 12/07/2024), Helios Hanseatic Hospital Stralsund (06/09/2024, first patient entered: 14/03/2025). The requirement for separate ethics approvals from Charité University Medical Center Berlin, University Medical Center Regensburg and Schoen Hospital Neustadt was waived by the ethics committees of the three participating hospitals, on the basis that their review processes recognize and accept the existing ethics approval granted by the University of Luebeck (2020-20-225). University Medical Center Regensburg signed their waiver on 30/09/2021 and included their first patient 28/04/2025. Charité University Medical Center Berlin became a part of the study on 04/12/2024 and Schoen Hospital Neustadt joined on 07/09/2022. Their first patient was included on the 09/12/2024 (Berlin) and on the 25/10/2022 (Neustadt), respectively. Patients participating in the study are informed beforehand and give written consent for their data to be collected. The end of the study after its 20-year period is the 31/09/2041 for all centers.

Study design

The GerPaCyst Study is a prospective, multicenter, interdisciplinary, longitudinal registry study. A formal protocol detailing aims and regulations for data safety and publication has been uploaded to https://drks.de/search/en/trial/DRKS00025927 and can be accessed publicly. This manuscript also provides a concise form to fulfill standards for documentation of registries and follows the Standard protocol items: recommendations for clinical trials (SPIRIT) guidelines [15].

Registry organization

The GerPaCyst-registry is headed by the Steering Committee including KCH, MD, RJ, SK, CM and UFW. The registry is physically located at the University of Luebeck, Germany. Maintenance is secured by the IT-Service Center Luebeck.

Participants, eligibility criteria and outcomes

Patients presenting at the respective participating centers including academic and academic-affiliated hospitals, with a presumed or confirmed pancreatic cyst (IPMN, MCN, SPPT, or SCN) will be evaluated for participating in the study. The diagnosis of a presumed pancreatic cysts will be established by the treating physician in accordance with the European evidence-based guidelines on pancreatic cystic neoplasms, the international, Sendai, Fukuoka criteria and the updated Kyoto guidelines [5,16–18], which defines the clinical, imaging and laboratory features required for case inclusion. Diagnosis criteria include the following:

• Main-Duct IPMN is suspected for segmental or diffuse dilation of the main pancreatic duct (MPD) of >5 mm without other causes of obstruction.
• Any cyst above or equal 1 cm should be checked by CT or MRI with MRCP to check for high-risk stigmata, including enhanced solid component and main pancreatic duct size of more than 1 cm, or worrisome features”, including cyst over or equal 3 cm, thickened enhanced cyst walls, non-enhanced mural nodules, MPD size 5–9 mm, abrupt change in MPD caliber with distal pancreatic atrophy and lymphadenopathy
• All cysts over 3 cms or with worrisome features or high-risk stigmata” should undergo EUS. MDCT and MRCP are most useful for distinguishing BD-IPMN from other cysts by showing multiplicity and a connection to the MPD. Cyst fluid is optional but can help to distinguish a BD-IPMN from a small oligocystic SCN with CEA determination.
• The distinction between main, mixed or branch-duct type IPMN should be made upon imaging. The pathological classification can only be done postoperatively.
• Neoplastic transformation will be separated into low-grade and high-grade dysplasia, as well as invasive cancer.
• MCN should be restricted to neoplasms exhibiting ovarian-type stroma.
• SPN/SPPTs are well-circumscribed mass with calcification, peripheral capsule, internal blood products, and lack of biliary/pancreatic ductal obstruction) on computed tomography and MRI are highly suggestive of the diagnosis of SPN, particularly when visualized in young female patients.

Imaging will be subjected to review at the University Medical Center Schleswig-Holstein, Campus Luebeck upon study inclusion for the prospective risk stratification model. For future research questions, MR and CT images will be reviewed at random to ensure quality of entered data. Discrepancies will be recorded and used to calculate inter-observer variability.

These are the inclusion and exclusion criteria:

1. Inclusion criteria:
   • Presumed pancreatic cyst of the following entities (IPMN, MCN, SPPT, SCN)
   • At least 18 years old
2. Exclusion criteria:
   • Confirmed pseudocyst or diagnosis other than IPMN, MCN, SPPT, SCN
   • Less than 18 years old

Suitable patients will be informed about the research project by the treating physician. The written informed consent will be obtained by each patient. The treatment interventions will not be altered by the study.

If the patient provides consent to participate in GerPaCyst, data entry will take place at the time of treatment initiation (surgery initiation of non-surgical surveillance) and every 6–12 months thereafter, for a period of up to 20 years. Participating centers complete an electronic Case Report Form (eCRF). Study data is collected and managed using REDCap electronic data capture tools hosted at the University of Luebeck, Germany [19,20] (Variables see S1 Table). Changes in patient characteristics are entered during patient follow-up. In addition, patient biobank samples (blood, cyst fluid, urine, tissue) will be stored upon availability.

The primary outcome of this trial is the development of high-grade dysplasia/invasive cancer during follow-up or the absence of it. The secondary outcomes are death, quality of life measured by EQ-5D-5L questionnaire, end-of-follow up and perioperative characteristics if applicable such as complications, length of hospital stay and clavien-dindo classification.

Initiation algorithm

The study is still welcoming new participating centers from all over the world. Interested centers can contact: kimchristin.honselmann@uksh.de or redcap.lachi@uni-luebeck.de. Once the study protocol has been reviewed by the ethics committee of the new center, and the contract has been approved by both legal departments, the participating physicians will receive proper training for patient inclusion in the REDCap based database.

Participant timelines and follow-up schedules

Only after the patient has given written informed consent may the treating physician register the patient in the database. After patient registration, mandatory baseline data on premedical history, data on the disease, imaging and diagnostic modalities, treatment, and treatment outcome (e.g., post-op complications) are documented. In addition, the date of treatment initiation, patient number, and basic patient-specific information are documented (e.g., gender or month and year of birth). The study’s non-interventional design does not regulate modalities or intervals between examinations and thus does not intervene in routine clinical practice. Patients will be followed when presenting to the participating institutions or by telephone interview every 6–12 months. A detailed patient schedule is shown (Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. SPIRIT schedule of enrollment, interventions, and assessment.

https://doi.org/10.1371/journal.pone.0335809.g001

Patient rights and data safety

Written informed consent for participation in the registry is obtained prior to inclusion at every center from all patients. Patient data are entered and stored in pseudonymized form. Institutional review boards have been obtained from every participating center. A data safety manual can be obtained from the senior author (KCH). The GerPaCyst platform was initiated and enabled for data collection in pancreatic cysts in 2021 (Ethics Committee of the University of Lübeck Az. 20−225 and 2024–265_1).

Assessment of data

After browser-based data entry, the data are automatically checked for completeness and internal plausibility by software-based internal rules. Missing or implausible items create warning messages. The data is usually entered after seeing the patient in the outpatient clinic of the respective institution. Missing data will be handled according to the intended analysis. There is currently no external monitoring for the GerPaCyst Registry. Data are collected prospectively via standardized physician-completed questionnaires. Diagnosis will be made according to the standardized criteria outlines in the aforementioned guidelines. Regular quality control is performed by semiannual study meetings of the entire study group. Through these periodic audits harmonized data collection is ensured, combined with the use of the electronic database that automatically identifies and flags incomplete case forms. The GerPaCyst steering committee recognizes that some methodological aspects may require refinement during implementation. Any such changes will be documented and reported. Specifically, laboratory analyses and parameters used in risk models, as well as imaging parameters, may be subject to change over the course of the study.

For building a multi-omics risk stratification model for pancreatic cyst progression, we will utilize deep learning methods. First, we will ensure data processing and integration and preprocess data by normalization (proteomics, genomics, metabolomics, etc.) and harmonization. Data will further be integrated by feature selection and dimensions will be reduced. Then deep learning will be integrated such as neuronal networks. These calculations will be performed by designated bioinformaticians.

Database structure

GerPaCyst is based on a web-based REDCap data entry system for the systematic collection of patient data, which was developed by Vanderbilt University [19]. Data entry is paperless and performed directly on-site at the respective medical facility (clinic/practice) via an Internet browser. The REDCap implementation meets all requirements of Good Clinical Practice (ICH-GCP) as well as European data safety regulations. Corresponding certificates for data security are available. The system is operated on a server of the computer center of the University of Lübeck and is thus subject to strict access control. Data is collected in pseudonymized form. Pseudonymization keys are kept separately from the database at the participating institutions.

Duration of observation per patient is up to a maximum of 20 years or until the occurrence of an event defined as end of follow-up or death. The study start was the fourth quarter of 2021 and is still ongoing. Centers are still being initialized at the moment.

Entered variables include baseline characteristics such as data on comorbidities like arterial hypertension, diabetes mellitus, chronic obstructive lung disease, medical history of cancer, family history and past medical history. Also, data on perioperative characteristics, if applicable, will be collected. Imaging modalities on the most prominent cysts will be collected at each follow-up. We will perform a quality-of-life questionnaire (EQ-5D-5L) at enrollment and at end of follow-up or surgery [21]. Cyst fluid analysis/molecular characteristics will be collected, if available (S1 Table).

Statistics

The data required for the respective research question are compiled into an anonymized project-specific evaluation data set with special consideration of the resulting re-identification risk, if any, and made available for the research project. The GerPaCyst study group decides on the provision of the evaluation data sets for scientific evaluation.

For the development of risk models, we aim to utilize neuronal networks with elastic-net regression using R. ROC-Curves will be built for candidate risk variables. According to our and the experience of others, sample size should be ten to 20 times higher than the number of trainable parameters. As of now, we are planning to include 157 variables plus imaging and biomarker profiles (about 40 different parameters). Our sample size should therefore be at least 4000 patients (200x20). To achieve adequate participant enrolment, this study was set up as an international multicenter study. Descriptive data will be evaluated with R or SPSS according to the data characteristics. Nominal data will be evaluated with frequencies and percentages and differences with Chi-square test or the Kaplan-Meyer methods if pertaining to survival. Continuous data will be evaluated by median and interquartile range, differences between two groups will be calculated by Mann-Whitney U Test. Kaplan-Meyer approximation will be used for survival analyses. P-value will be set at p < 0.05.

The further exact analyses will be defined in a Statistical Analysis Plan (SAP) prior to the corresponding evaluation. However, to ensure appropriate analyses, the nature of the data must be considered. For this reason, descriptive analyses will be conducted, from which new hypotheses and questions will arise. Potential systematic biases must be identified and accounted for in the analyses. To account for interobserver variability in radiological data sets, each time a scientific question is raised, it is followed by a review of the imaging data supplied. Other important decisions include defining subgroups of interest (e.g., patient groups that are usually excluded from RCTs) or pooling similar therapies (e.g., when there are only a few patients in individual therapies) and dealing with missing values in registry data. Some of the questions result in periodic evaluations. Other questions arise due to the ongoing, observational process only during the study, which then result in supplementary SAPs. Answering questions that arise only in the course of the study is one of the great advantages (and difficulties) of registry studies [22].

Only the ITSC of the University of Lübeck and the leaders of the GerPaCyst study group or persons appointed by the study group, such as staff of the Chir-NET office of the University Hospital Schleswig-Holstein, have access to the complete data set.

The GerPaCyst registry has been established to be used together with the StuDoQIPancreas Registry, if necessary and for facilitating data use and transfer within Germany, but also internationally. Accordingly, many items have been adapted from the StuDoQIPancreas registry [23].

Dissemination policy

Trial results will be communicated with all participating centers on a regular basis including half-yearly meeting sessions at the German Pancreas Club and the German Visceral Medicine meetings (AG Pankreas DGVS). In addition, 4-monthly group meetings will be held virtually in order to communicate possible protocol changes or interim-results. Publication will be decided upon by the steering committee and author inclusion will be based on patient inclusion: up to 10 patients per year: 1 author from each center, 10–20 patients per year: 2 authors per center, etc. Requests for data analysis of the whole cohort can be send to the steering committee for evaluation. The full protocol has been uploaded to the German clinical trial registry (www.drks.de). Participant-level dataset and statistical codes can be obtained from KCH.

Biological specimens

Blood samples or cyst (fluid) biopsies or surgical specimens will be collected at the participating centers, if available and will possibly be used for genetic and molecular analysis in ancillary studies in the future.

Appendices

Informed consent (Currently in German).

Trial status

11 university hospitals throughout Germany are currently participating in this study. The first patient was entered in December 2021 and the study is ongoing. It is an interdisciplinary cooperation of surgeons. radiologists and gastroenterologists. At least 5000 patients are planned to be recruited from gastroenterology and surgical departments. Baseline characteristics such as metabolic/endocrine diseases, risk factors and family history, as well as pancreatic cyst data including MPD diameter, enhanced mural nodules, calcifications and the quality-of-life EQ-5 questionnaire will be assessed during index visit and follow-up. Biospecimens like blood, cyst fluid and cyst cytology will be collected, if needed.