We read with interest the recent correspondence of Germer et al.1 and Chevaliez et al.2 regarding the quantification of genotype 4 hepatitis C virus (HCV) RNA by the COBAS AmpliPrep/COBAS TaqMan HCV Test (CAP/CTM) (Roche Molecular Systems Inc., Branchburg, NJ). Several publications evaluating the quantification of genotype 4 serum samples have led to conflicting reports.3-9
We evaluated the correlation between viral load results in the serum samples of 75 pretreatment patients infected with genotype 4 and the impact of nucleotide (nt) polymorphism at nt 145 and nt 165 on viral load quantification. HCV RNA measurements were performed with the Versant HCV 3.0 Assay (branched DNA) (Siemens Healthcare Diagnostics Inc., Saint Denis, France); the CAP/CTM test; and the Abbott m2000sp extraction/m2000rt amplification system (ART) (Abbott Laboratories Inc.) and the COBAS AmpliPrep/COBAS TaqMan HCV Test (Roche Molecular Systems).
HCV genotypes were identified using the TruGene HCV 5′NC genotyping kit (Siemens Healthcare Diagnostics Inc.). HCV subtyping was performed in the NS5 B nonstructural region of the HCV genome with the Open Gene Thermo sequenase fluorescent-labeled primer cycle sequencing kit (Siemens Healthcare Diagnostics Inc.).10
The mean viral loads for the 75 serum samples (HCV genotype 4a, n = 36; 4c, n = 1; 4d, n = 16; 4e, n = 10; 4f, n = 4; 4h, n = 4; 4i, n = 2; and 4l, n = 2) were: 5.300 ± 0.751, 5.334 ± 0.941, and 5.419 ± 0.820 with the branched DNA, CAP/CTM, and ART tests, respectively (all values are not significant). Our results showed similar quantification levels for HCV genotype 4 subtype, no matter which assay was used. These data are in agreement with those reported by Germer et al. on a cohort of 100 clinical genotype 4 samples.1
In our 75 patients, HCV 5′ noncoding region sequences revealed no sequence containing a G-to-A substitution at nt 145, which was reported to be associated with failure of CAP/CTM by Chevaliez et al.2; three sequences contained an A-to-T substitution at nt 165, which was previously associated with under-quantification by CAP/CTM.2 Two of these three substitutions at nt 165 yielded under-quantification of 0.5 log10 IU/mL and 0.988 log10 IU/mL with CAP/CTM. These results are in accordance with those of Germer et al.1 and confirm that substitutions at nt 145 and 165 are very rare. Thus, one would be unlikely to encounter failure of CAP/CTM to detect HCV genotype 4 strains not only in U.S. samples but also in European samples.
In conclusion, our results show similar quantification levels for the different HCV genotype 4 subtypes, irrespective of whichever assay was used, and that a G-to-A substitution at nt 145 and an A-to-T substitution at nt 165 are very rare, confirming that one would be unlikely to encounter failure of CAP/CTM to detect HCV genotype 4 strains found in the United States and in Europe. Nevertheless, it is important to consistently use the same HCV RNA assay throughout patient treatment follow-up.