Background and objectives: Since July 1 1999, four laboratories in the Netherlands have been routinely screening plasma minipools for the release of labile blood components utilizing hepatitis C virus nucleic acid amplification technology (HCV NAT). This report describes the performance evaluation of the HCV NAT method and the quality control results obtained during 6 months of routine screening.
Materials and methods: Plasma minipools of 48 donations were prepared on a Tecan Genesis robot. HCV RNA was isolated from 2 ml of plasma by using the NucliSens Extractor and amplified and detected with the Cobas HCV Amplicor 2.0 test system. For validation of the test system the laboratories used viral quality control (VQC) reagents of CLB.
Results: Initial robustness experiments demonstrated consistent detection of PeliSpy HCV RNA samples of 140 genome equivalents/ml (geq/ml) in each station of the installed Nuclisens Extractors. Further 'stress' tests with a highly viraemic sample of approximately 5 x 10(6) geq/ml did not contaminate negative samples processed on all Extractor stations in subsequent runs. In the validation period prior to July 1999, 1021 pools were tested with the following performance characteristics: 0.1%, initially false reactive; 0.89%, failure of internal control detection; 0.97%, no eluate generated by the Extractor; and 100% reactivity of the PeliSpy 140 geq/ml control in 176 Extractor runs and a 98% reactivity rate of the PeliSpy 38 geq/ml control in 102 test runs. By testing the PeliCheck HCV RNA genotype 1 dilution panels 49 times, an overall 95% detection limit of 30 geq/ml ( approximately 8 IU/ml) and a 50% detection limit of 5 geq/ml was found by the four laboratories. In the first 6 months of routine screening, the minimum requirement for invalid results (2%) was exceeded with some batches of silica and NucliSens Extractor cartridges. From November 1999 to February 2000, the manufacturer (Organon Teknika) improved the protocol for silica absorption of the Nuclisens Extractor -- the cartridge design as well as the software of the Extractor. During the next 6 months of observation in 2000, the percentages of false initial reactives and invalids were 0.05% and 1.4%, respectively, in 8962 pools tested. Of these invalid results, 0.74% and 0.66% were caused by Extractor failure and negative internal control signals, respectively. The PeliSpy HCV RNA 'stop or go' run control of 140 geq/ml was 100% reactive, but invalid in 16/1375 (1.2%) of cases. The PeliSpy run control of 38 geq/ml for monitoring sensitivity of reagent batches was reactive in 95% of 123 samples tested.
Conclusions: Each of the four HCV NAT laboratories in the Netherlands have achieved similar detection limits that are well below the sensitivity requirements of the regulatory bodies. After improvement of the NucliSens Extractor procedure, the robustness of the test system has proved to be acceptable for routine screening and timely release of all labile blood components.