| 262P | |
PROCEDURE FOR SUSTAINING RULES BASED DRUG RESISTANCE INTERPRETATION ALGORITHMS
CL Reid1, DL Winslow1, K Smith1,
JD Baxter2, F Brun-Vézinet3, B Clotet4, RM
Grant5, C Loveday6, TC Merigan7, W O’Brien8,
MA Wainberg9, JM Schapiro7
1Visible Genetics, Toronto, Ontario; 2Cooper Hospital/RWJ Medical
School, Camden, New Jersey; 3Hôpital Bichat-Caude Bernard, Paris,
France; 4Hosp. Universitari Germans Trias i Pujol, Barcelona, Spain;
5Gladstone/UCSF, San Francisco, California; 6Royal Free/University
College Medical School, London, United Kingdom; 7Stanford University,
Palo Alto, California; 8University of Texas Medical Branch, Galveston,
Texas; 9McGill University AIDS Center/Jewish General Hospital, Montreal,
Quebec
Objectives: Genotypic drug resistance assays generate mutation lists
that require interpretation to be clinically useful. HIV resistance is a rapidly
evolving field characterized by frequent presentation of new data. There is
a need to define a process by which interpretation algorithms are established
and new data routinely incorporated. The Visible Genetics (VGI) Consensus Panel,
an academic group convened by VGI, developed and maintains a rules-based algorithm
for the TRUGENEÔ HIV-1 Genotyping Test.
VGI and the Panel have jointly developed a successful process of rules preparation
and revision. Methods: The Panel created the first rules set in 1999 and revised
the set biannually thereafter. Procedural elements essential to the validity
and continued clinical relevance of the algorithm were formalized in a detailed
document consistent with medical device industry Design Control standards.
Results: Key elements identified were:
| • | Comprehensive knowledge of drug resistance data and ongoing resistance research as Panel member qualifications |
| • | 8 member quorum for each revision meeting including scientific chairman |
| • | Biannual revision meetings supported by follow up teleconferences |
| • | Comprehensive review of all newly presented data |
| • | Extensive Panel discussion and consensus agreement regarding all rules revisions independent of sponsor influence |
| • | References identified for all rules and revisions maintained in a comprehensive database |
| • | Evidence based system for ranking rules |
| • | Specific quality control and rules set validation method |
| • | Documentation of Panel decisions and significant discussions |
| • | Coordination of process by Panel administrator |
The fourth of five rules sets was FDA cleared in 2001 as part of the TRUGENE
HIV-1 Genotyping Kit and the OpenGeneÔ
DNA Sequencing System, based upon submission of procedural documents and specific
supporting references.
Conclusions: Clinically relevant, up-to-date rules-based interpretation
algorithms require an established procedure involving independent scientific
expertise to incorporate the most relevant new data in a timely and accurate
manner.