Development and verification of an automated sample processing protocol for quantitation of human immunodeficiency virus type 1 RNA in plasma

Abstract

We developed and verified an automated sample processing protocol for use with the AMPLICOR HIV-1 MONITOR test, version 1.5 (Roche Diagnostics, Indianapolis, Ind.). The automated method uses the MagNA Pure LC instrument and total nucleic acid reagents (Roche Applied Science, Indianapolis, Ind.) to extract human immunodeficiency virus type 1 (HIV-1) RNA from plasma specimens. We compared the HIV-1 load results for a dilution series (1 to 5 nominal log(10) copies/ml) and 175 clinical specimens processed by the automated method to those for the same samples processed by the manual methods specified by the manufacturer. The sensitivity, dynamic range, and precision of the viral load assay obtained by automated processing of specimens were similar to those obtained by an ultrasensitive manual processing method. The results were highly correlated (R(2), 0.95), and were in close agreement, with a mean difference of 0.09 log(10) (standard deviation, 0.292). The limits of agreement were +/-0.58 log(10) for results for samples processed by both the manual and the automated methods. These performance characteristics were achieved with a smaller sample volume (200 versus 500 microl) and without a high-speed centrifugation step and required only 15 min of labor for a batch of 32 samples. In conclusion, the automated sample preparation protocol can replace both the standard and the ultrasensitive manual methods used with the AMPLICOR HIV-1 MONITOR test and can substantially reduce the labor associated with this test.

FIG. 1.

Plot of measured against expected log₁₀ number of HIV-1 copies per milliliter for specimens processed by the MPLC protocol. Each point represents the mean for eight replicates. The equation for the linear regression line was y = 0.86x + 0.73 (R², 0.996).

FIG. 2.

Correlation of viral load results for all clinical specimens processed in parallel by the MPLC and the manual protocols. ▪, ultrasensitive manual method; ▴, standard manual method; solid line, trend line for values determined by the ultrasensitive manual and MPLC protocols (R², 0.913); dashed line, trend line for values determined by the standard manual and MPLC protocols (R², 0.863).

FIG. 3.

Difference in log₁₀ number of HIV-1 copies per milliliter against average log₁₀ number of HIV-1 copies per milliliter for specimens processed manually and by the MPLC protocol. Solid line, mean difference (0.086 log₁₀); dashed lines, ±2 standard deviations (0.568 log₁₀); ▪, ultrasensitive manual method; ▴, standard manual method.

FIG.4.

Correlation of viral load results for clinical specimens with viral load values between 1.7 and 5 log₁₀ copies/ml processed in parallel by the MPLC and the manual protocols. The equation for the linear regression line was y = 1.082x + 0.354 (R², 0.945).