No abstract

Since 1991, a nationwide histopathology and cytopathology network and archive is in operation in The Netherlands under the name PALGA, encompassing all sixty-four pathology laboratories in The Netherlands. The overall system comprises decentralized systems at the participating laboratories, a central databank, and a dedicated communication and information exchange tool. Excerpts of all histopathology and cytopathology reports are generated automatically at the participating laboratories and transferred to the central databank. Both the decentralized systems and the central system perform checks on the quality and completeness of excerpts. Currently, about 42 million records on almost 10 million patients are stored in the central databank. Each excerpt contains patient identifiers, including demographic data and the so-called PALGA diagnosis. The latter is structured along five classification axes: topography, morphology, function, procedure, and diseases. All data transfer and communication occurs electronically with encryption of patient and laboratory identifiers. All excerpts are continuously available to all participating pathology laboratories, thus contributing to the quality of daily patient care. In addition, external parties may obtain permission to use data from the PALGA system, either on an ongoing basis or on the basis of a specific permission. Annually, 40 to 60 applications for permission to use PALGA data are submitted. Among external users are the Dutch cancer registry, population-based screening programs for cancer of the uterine cervix and breast cancer in The Netherlands, and individual investigators addressing a range of research questions. Many scientific papers and theses incorporating PALGA data have been published already. In conclusion, the PALGA system is a unique system that requires a minimal effort on the part of the participating laboratories, while providing them a powerful tool in their daily practices.

These preliminary findings suggest an association between silicone breast prostheses and ALCL, although the absolute risk is exceedingly low due to the rare occurrence of ALCL of the breast (11 cases in The Netherlands in 17 years). These findings require confirmation in other studies.

Gastrointestinal stromal tumours (GIST) are identified by their specific morphology added with immunohistochemical staining with anti-CD117, the phenotypic marker for GIST in the majority of cases. In this study the reported incidence of GIST and GIST-like tumours before and after the routine availability of the specific diagnostic marker CD117 antigen was investigated. All patients with GIST or GIST-like tumours were selected from PALGA, a nation-wide network and registry of histo- and cytopathology in the Netherlands, to calculate the incidence in 1995 and longitudinally over time between 1998 and 2003. Pathological reports were retrieved to assess the type of immunostaining used and to assess the risk category for malignant behaviour according to the recently published consensus criteria. The annual incidence of GIST as distilled from the national pathology registry increased from 2.1 per million inhabitants in 1995 to 12.7 per million inhabitants in 2003. The incidence of GIST-like tumours decreased from 17.6 per million inhabitants in 1995 to 12.7 per million inhabitants in 2003. The incidences were stable from 2000 onwards. Additional analysis in 2003 indicated that more than 90% of the GIST tested was CD117 positive, compared to only 4% of the GIST-like tumours. Almost 50% of the GIST was considered to be at high-risk for malignant behaviour, according to the consensus criteria. The increased incidence of GIST 1995-2003 is related to increased understanding of GIST pathobiology and the routine availability of the diagnostic immunohistochemical antibody directed against the CD117 antigen.