Babette E Becherer1, Mintsje de Boer1, Pauline E R Spronk1, Annette H Bruggink1, Jan Paul de Boer1, Flora E van Leeuwen1, Marc A M Mureau1, René R J W van der Hulst1, Daphne de Jong1, Hinne A Rakhorst1. 1. From the Department of Plastic and Reconstructive Surgery, Erasmus MC Cancer Institute University Medical Center Rotterdam; the Dutch Institute for Clinical Auditing, Dutch Breast Implant Registry; the Department of Plastic, Reconstructive, and Hand Surgery, Maastricht University Medical Centre; the Dutch Nationwide Network and Registry of Histo- and Cytopathology; the Departments of Medical Oncology and Epidemiology and Biostatistics, Netherlands Cancer Institute; the Department of Pathology, Amsterdam University Medical Center, VU University Medical Centre; and the Department of Plastic, Reconstructive, and Hand Surgery, Ziekenhuis Groep Twente/Medisch Spectrum Twente.
Abstract
BACKGROUND: The Dutch Breast Implant Registry (DBIR) was established in April of 2015 and currently contains information on 38,000 implants in 18,000 women. As a clinical registry, it evaluates the quality of breast implant surgery, including adverse events such as breast implant-associated (BIA) anaplastic large cell lymphoma (ALCL). To examine the efficacy of the DBIR, the capture rate of BIA-ALCL was compared to the registration of BIA-ALCL in the Dutch Nationwide Network and Registry of Histo- and Cytopathology (PALGA) as a gold standard, in combination with matching these databases to obtain complementary information. METHODS: All BIA-ALCL patients diagnosed and registered in The Netherlands in 2016 and 2017 were identified separately in the PALGA and DBIR databases. In addition, both databases were matched using indirect key identifiers. Pathologic information from the PALGA and clinical and device characteristics from the DBIR were obtained for all patients. RESULTS: Matching of both databases gave a capture rate of BIA-ALCL in the DBIR of 100 percent (n = 6) in 2016 and 70 percent (n = 7) in 2017. In total, 17 patients were identified in the PALGA, of which 14 patients were also identified in the DBIR; three patients were not registered; and 10 patients were registered false-positive. Of all confirmed patients, symptoms, staging results, treatment, and implant information were registered. CONCLUSIONS: Currently, the DBIR contains 2 full registration years and captures most of the BIA-ALCL patients despite overestimation. Therefore, pathology confirmation remains essential. By matching these databases, complementary clinical and implant information could be retrieved, establishing the DBIR as an essential postmarketing surveillance system for health risk assessments.
BACKGROUND: The Dutch Breast Implant Registry (DBIR) was established in April of 2015 and currently contains information on 38,000 implants in 18,000 women. As a clinical registry, it evaluates the quality of breast implant surgery, including adverse events such as breast implant-associated (BIA) anaplastic large cell lymphoma (ALCL). To examine the efficacy of the DBIR, the capture rate of BIA-ALCL was compared to the registration of BIA-ALCL in the Dutch Nationwide Network and Registry of Histo- and Cytopathology (PALGA) as a gold standard, in combination with matching these databases to obtain complementary information. METHODS: All BIA-ALCLpatients diagnosed and registered in The Netherlands in 2016 and 2017 were identified separately in the PALGA and DBIR databases. In addition, both databases were matched using indirect key identifiers. Pathologic information from the PALGA and clinical and device characteristics from the DBIR were obtained for all patients. RESULTS: Matching of both databases gave a capture rate of BIA-ALCL in the DBIR of 100 percent (n = 6) in 2016 and 70 percent (n = 7) in 2017. In total, 17 patients were identified in the PALGA, of which 14 patients were also identified in the DBIR; three patients were not registered; and 10 patients were registered false-positive. Of all confirmed patients, symptoms, staging results, treatment, and implant information were registered. CONCLUSIONS: Currently, the DBIR contains 2 full registration years and captures most of the BIA-ALCLpatients despite overestimation. Therefore, pathology confirmation remains essential. By matching these databases, complementary clinical and implant information could be retrieved, establishing the DBIR as an essential postmarketing surveillance system for health risk assessments.
Authors: Elaine S Jaffe; Binita S Ashar; Mark W Clemens; Andrew L Feldman; Philippe Gaulard; Roberto N Miranda; Aliyah R Sohani; Timothy Stenzel; Sung W Yoon Journal: J Clin Oncol Date: 2020-02-11 Impact factor: 44.544