LETTER

Top Letter References

Recently, Harasawa and Tomiyama (1) reported the detection of pestivirus RNA in human vaccines. The origin of the contamination was unknown, but the vaccines had been prepared in cell cultures supplemented with bovine fetal calf serum (FCS), which can be contaminated with bovine viral diarrhea virus (BVDV).

We used a closed one-tube reverse transcription (RT)-nested PCR method to attempt the amplification of pestivirus RNA from 30 human vaccines produced by European companies. Six were inactivated vaccines made in embryonated eggs and contained purified influenza viruses. Twenty-four were live unpurified, cell culture-derived vaccines and contained polio, measles, mumps, or rubella viruses (six samples of each). Six control samples were also provided and contained 1% calf seurm or 1% FCS (three samples of each). The 36 samples were tested blind in nine batches. Each batch of testing incorporated a positive-control sample (BVDV) and seven negative-control samples (water).

An RNA isolation procedure was carried out on all 108 samples and controls. This involved addition of 800 µl of Trizol (Sigma) to 200 µl of sample or control, followed by processing according to the manufacturer's recommendation. To minimize false-positive reactions, we used a closed one-tube RT-nested PCR method developed in our laboratory. Half of the extracted RNA was added to a tube containing RT and first PCR mixtures, along with outer panpestivirus primers V324 and V326. Fresh Taq DNA polymerase and inner panpestivirus primers A11 and A14 were present in the tube lid in a dry form. These components were added to the second (nested) PCR mixture (30 cycles) by inversion of the tube after RT and the first PCR (20 cycles). Amplified products were examined by agarose gel electrophoresis and sequencing. Details of this procedure will be described elsewhere and show that the method is more sensitive than virus isolation for detection of pestiviruses in clinical samples (2).

The method detected BVDV in two of three control samples containing 1% FCS and in one of three samples of 1% calf serum. Sequencing two amplicons showed that a different pestivirus had been detected in each sample type. Both products had been declared by the supplier to be pestivirus free. Other negative and positive controls gave the expected results. Pestivirus RNA was not detected in any of the vaccine samples tested. Four vaccine samples that gave suspicious results were 50× concentrated by ultracentrifugation. Repeat testing and sequencing of amplified products confirmed only nonspecific amplifications. Our results demonstrate that the frequency of contamination of human vaccines by pestiviruses may be lower than has been reported previously.