Campbell WF, Tau KR, Michel-Treil VP, Hutchinson E; International Conference on AIDS.
Int Conf AIDS. 2000 Jul 9-14; 13: abstract no. MoPeA2113.
W.F. Campbell, Covance Central Laboratory Services, 8211 SciCor Drive, Indianapolis, IN 46214, United States, Tel.: 317-273-7805, Fax: 317-273-7990, E-mail: bill.campbell@covance.com
Background: HIV-1 plasma Viral Load determination is the marker of choice to assess efficacy of antiretroviral therapy in AIDS clinical trials. Several PCR method formats have be developed to extend the range of quantifiable RNA copy numbers (Roche Amplicor Standard and Ultrasensitive) and to identify non-B HIV-1 subtypes (Roche Amplicor Version 1.5) which are being isolated with increasing frequencies in Europe. The inherent analytical variation of PCR is related to the amplification process itself. To provide combinable data in global studies, it is imperative that technical variation be no greater between two testing sites than within a single testing site. Methods: Three different levels of controls (plasma pools prepared by Boston Biomedica, Inc.) were analyzed using the Roche Amplicor Version 1.0 and 1.5 by our laboratories in Geneva and in Indianapolis. The two lowest levels of controls were analyzed in replicate using with the Roche Amplicor Ultrasensitive assay and the highest level of control was analyzed using the Roche Amplicor Standard assay. To establish equivalency of testing in the two facilities, accuracy was assessed by comparing the ratio of means while precision was assessed by comparing the ratio of standard deviations. Results: For the Level 1 control using the 1.0 Ultrasensitive assay, Geneva and Indianapolis generated means of 155 C/mL and 198 C/mL, respectively while the standard deviations were 64 and 73, respectively. Using the 1.5 Ultrasensititive assay, the means for Geneva and Indianapolis were 126 C/mL and 152 C/mL, respectively and the standard deviations were 49 and 76, respectively. For the Level 2 control using the Ultrasensitive 1.0 assay, Geneva and Indianapolis generated means of 1450 C/mL and 1515 C/mL, respectively while the standard deviations were 378 and 421, respectively. Using the 1.5 Ultrasensititive assay, the means for Geneva and Indianapolis were 1038 C/mL and 1500 C/mL, respectively and the standard deviations were 357 and 446 respectively. For the Level 3 control using the Standard 1.0 assay, Geneva and Indianapolis generated means of 17440 C/mL and 12897 C/mL, respectively while the standard deviations were 3665 and 2151, respectively. Using the 1.5 Standard assay, the means for Geneva and Indianapolis were 17831 C/mL and 15736 C/mL, respectively and the standard deviations were 6134 and 5955, respectively. Correlations between laboratories in Indianapolis and Cape Town and between Geneva and Cape Town also were acceptable. Conclusions: The ratio of means was well within the acceptable limits defined by Roche (threefold variation) for all tests compared. Similarly, the ratio of standard deviations was well within the acceptable limits for all tests compared. Thus, accuracy and precision were equivalent in the three testing facilities indicating that data generated are combinable.
Publication Types:
Keywords:
- Acquired Immunodeficiency Syndrome
- Boston
- Clinical Trials as Topic
- Europe
- Evaluation Studies
- HIV-1
- Laboratories
- Laboratory Techniques and Procedures
- Polymerase Chain Reaction
- Research Design
- South Africa
- Viral Load
- organization & administration
- virology
Other ID:
UI: 102237660
From Meeting Abstracts