T R A N S F U S I O N C O M P L I C A T I O N S

Mini-pool screening by nucleic acid testing for hepatitis B virus, hepatitis C virus, and HIM preliminary results

M.S. Cardoso, K Koerner, and B. Kubanek

BACKGROUND: The purpose of this study was to evaluate the feasibility of nucleic acid testing (NAT) of mini-pools as a blood donation screening test. STUDY DESIGN AND METHODS: The stepwise imple- mentation of NAT of mini-pools began in January 1997. Since March 1997, all blood donations collected by the German Red Cross Blood Transfusion Service of Baden-Wurttemberg were tested for hepatitis B virus (HBV), hepatitis C virus (HCV), and HIV nucleic acids. An extra barcoded serum sample is collected from each blood donor for NATbased screening, which is per- formed only on hepatitis B surface antigen-, anti-HCV-, anti-HIV-, and anti- Treponema pallidumseronegative do- nations. Samples are pooled to a maximum of 96. Posi- tive results are resolved through intersecting subpools (a chessboard design). NATbased screening does not include a virus concentration step before nucleic acid extraction. RESULTS: By the end of October 1997,331,783 dona- tions in 3,779 pools had been screened. As yet, no viremic but seronegative blood donor has been found for the three markers. CONCLUSION: It is feasible to incorporate NAT-based screening of mini-pools into the routine virus diagnostics of a large blood transfusion service. It remains to be de- termined whether screening blood donations by NAT will indeed increase the safety of blood supply.

ABBREVIATIONS: HBV = hepatitis B virus; HCV = hepatitis C virus; NAT = nucleic acid testing; PCR = polymerase chain reac- tion.

From the German Red Cross Blood Transfusion Service of Baden-Wiirttemberg and the Department of Transfusion Medi- cine, University of Ulm, Ulm, Germany.

ceived February 24,1998, and accepted April 14,1998. Received for publication November 24, 1997; revision re-

TRANSFUSION 1998;38:905-907.

n Germany, the regulating agency for blood compo- nents, the Paul Ehrlich Institute, announced in March 1997 its intention to introduce the nucleic acid test- I ing (NAT)-based screening of blood donations for

hepatitis B virus (HBV), hepatitis B virus (HCV), and HIV DNAIRNA. This decision was based mainly on data re- ported by the German Red Cross Blood'Ikansfusion Service of North Rhine-Westphalia, which claimed at the end of 1996 to have detected 6 HBV-, 69 HCV-, and 0 HIV-viremic but -seronegative blood donors through polymerase chain reaction (PCR) screening of 650,403 blood donors.' Subse- quently, these results have been reanalyzed because of unconfirmed HCVPCR results (Schottstedt \5 presented at the National Hearing at the Paul Ehrlich Institute, Septem- ber 25, 1997).

The German Red Cross Blood 'Ikansfusion Service of Baden-Wurttemberg began in January 1997 the stepwise implementation of NAT-based screening for the three vi- ruses. The primary goal was to implement NAT-based screening as a releasing criterion for plasma components. Cellular components will be considered in a later phase.

Since March 1997, an average of 2000 blood donations have been screened daily. By the end of October 1997, 331,783 donations in 3,779 pools of a maximum of 96 do- nor samples per pool had been screened.

MATERIALS AND METHODS An extra barcoded serum sample is collected from each blood donor for NAT. Samples are transported at 20°C and stored at 4°C until the beginning of the testing (time lag, maximum of 18 hours). 'Itansport and storage conditions have been validated.2

NAT-based screening is performed only on hepatitis B surface antigen-, anti-HCV-, anti-HIV-, and anti- Deponema pallidum-seronegative donations. It does not include a vi- rus concentration step before nucleic acid extraction. All donations are first distributed into three identical barcoded microtiter plates, which are covered and stored as sample backup. After the conclusion of the standard serologic tests, the seronegative donations are selectively pooled (100 pL/ donation) to a maximum of 96 samples with a multi-

Volume 38, October 1998 TRANSFUSION 905

CARDOSO ET AL.

pipetter (Genesis 150, TECAN, Crailsheim, Germany). Commercial kits (COBAS HCVAmplicor, Roche Diagnostic Systems, Grenzach-Wyhlen, Germany; and AcuGen HBV and HIV kits, Biotronics, Lowell, MA) were used for the screening of pools of donations. The sensitivity of tests (single-sample testing) was determined to be 1000 copies per mL for HCV and 5000 copies per mL for HBV and H N with a sensitivity of 90 and 95 percent, respectively. Sensi- tivity was calculated as the numbers of PCR results divided by total number of tests with a positive sample.

The quality of laboratory performance was attested in 1997 by an HCV RNA proficiency study (VQC, CLB, Amsterdam, the Netherlands) and regularly by control pan- els for qualitative and quantitative analysis of HCVand HIV RNA (Instand, Diisseldorf, Germany). RNA isolation with kits (High Pure RNA Isolation Kit, Boehringer Mannheim, Mannheim, Germany; or QIAamp Viral RNA Kit, QIAGEN, Hilden, Germany) and amplification were performed ac- cording to the manufacturers’ instructions. A positive con- trol serum and a negative control serum were used as PCR controls for each run of the three assays.

In the COBAS HCVAmplicor test, the amplification of HCV-specific sequences occurs simultaneously with the amplification of an internal control RNA (in vitro RNA tran- script that is HCV irrelevant), which is added to each test sample at the RNA-solubilization step. In the AcuGen test, a generic amplification (plasmid DNA) follows the virus- specific sequence amplification to evaluate the quality of the sample.

A PCR-positive reaction will have to be confirmed in a secondary pooling scheme (a chessboard design). In this design, a reactive pool of 96 samples will be divided into 8 pools of 12 samples, each corresponding to a row of the second microtiter plate, and 12 pools of 8 samples, each corresponding to a column of the second microtiter plate.3 nYo reactive pools, one of 8 samples and another of 12 samples, indicate a positive donation. This finding has to be confirmed by individual PCR testing of the suspected donation. Data management is carried out by a modified software developed by the German Red Cross BloodTrans- fusion Service of North Rhine-Westphalia.

The sensitivity of the assays in pools of 96 samples was calculated to be 5 x lo5 copies per mL for HBVand HIVPCR- based screening and 1 x lo5 copies per mL for HCV PCR- based screening. In the meantime, various protocols to improve sensitivity through the concentration of virus par- ticles, particularly H W by centrifugation were investigated and will be implemented in routine testing in the f u t ~ r e . ~

RESULTS From January until November 1997, a total of331,783 blood donations were screened by NAT for HBV, HCX and HIV nucleic acids. These donations have been tested in pools of

a maximum of 96 samples, with the total at the end of Oc- tober reaching 3779 tested pools.

Preliminary results acquired by October 31, 1997, are as follows: the total of PCR-screened donations was 331,783; the number of HBV DNA-, HCV RNA-, and HIV RNA-posi- tive donations was 0; and the rate of false-positive results by HBV PCR was 2 percent; by HCV PCR was 0.6 percent, and by HIVPCR was 0.7 percent. False-positive results were those from pools that reacted in the NAT-based screening, but which could not be confirmed as positive in secondary testing (a chessboard design).

DISCUSSION Efforts to improve the sensitivity of NAT-based screening of blood donations for HBV, HCV, and HIV through the con- centration of virus particles by centrifugation have been ex- tensively investigated, particularly with regard to HCV A moderate increase in sensitivity can be ob~erved .~ A new protocol, including the concentration step, will be imple- mented in the future. Yet, even without this concentration step, an HCV window-period donation would have been detected because of the high titer of virus in serum at this t ime-p~in t .~ HCV RNA titers ranging from 3000 to 2 x lo6 RNA copies per mL, with a median value of 1.6 x lo5 RNA copies per mL, were observed by us in serum from 41 blood donors who seroconverted (data not shown).

Unspecific amplification leading to false-positive re- sults was observed in all three tests. To date, there has never been evidence of multiple crossover contamination in a series of samples, probably because of the strict mainte- nance of prevention measures.6 In HBV tests, the rate of false-positive results was significantly higher, which might be due to initial problems with the definition of the cutoff value.

In conclusion, our results suggest that the residual risk of transmitting HCV infection through blood donations is not higher than that calculated from seroconversion rates. In our donor population, of which over 90 percent are re- peat donors, the calculated residual risk is 1 in 330,0OO.’Re- sidual risks for HBVand HIV transmission are calculated to be 1 in 320,000 and 1 in 1.1 x lo6, respectively, and those levels fit well with the collected PCR data. Moreover, it be- came evident from the reported study that it is indeed fea- sible to incorporate NAT-based screening of mini-pools into the routine virus diagnostics of a relatively large blood transfusion service (approx. 450,000 donations/year). It re- mains to be determined whether the cost-benefit ratio of testing for the three viruses will justify such an enterprise, taking in account the extremely low residual risk of trans- mission and the question of whether screening blood do- nations by NATwill indeed increase the safety of blood sup- ply If the latter is the case, testing of single samples should be striven for in the future.

906 TRANSFUSION Volume 38, October 1998

MINI-POOLS AND NAT

REFERENCES 1. Schottstedt V, Tuma W, Biinger G, Lakenberg P PCR-Rou-

tine-Screening auf HBV-, HCV-, und HIV-1-Genom in einem groBen Blutspendedienst-Erfahrungen und erste Ergebnisse (abstract). Infusionsther Transfusionsmed 1996;23(S3):32. Cardoso MS, Koerner K, Kubanek B. [Experience of PCR testing in the routine blood banking of the DRK- Blutspendedienst Baden-Wiirttemberg] . Infusionsther Transfusionsmed 1998;25:116-20. Mortimer J. Intersecting pools and their potential applica- tion in testing donated blood for viral genomes. Vox Sang

Cardoso MS, Koerner K, Epple S, et al. Mini-pool testing by the polymerase chain reaction for viral nucleic acids: HCV concentration efforts (letter). Vox Sang 1998;74:262. Busch MP, Rawal BD, Nowicki M, et al. HCV RNA and ALT patterns in seroconverting transfusion recipients, and cor- relation with donor ALT and RNA (abstract). Transfusion

2.

3.

1997;73:93-6. 4.

5.

1997;37(Suppl):111S.

6. Kwok S, Higuchi R. Avoiding false positives with PCR. Na- ture 1989;339:237-8. Koerner K, Cardoso M, Dengler T, et al. Estimated risk of transmission of hepatitis C virus by blood transfusion. Vox Sang 1998;74:213-6.

7.

AUTHORS

Marcia S. Cardoso, Dr Biol Hum, Department of Transfusion Medicine, University of Ulm; and Head, Molecular Virus Diag- nostics Section, German Red Cross Blood Transfusion Service of Baden-Wiirttemberg, Helmholtzstrasse 10,89081 Ulm, Ger- many. [Reprint requests]

Department, German Red Cross Blood Transfusion Service of Baden-Wiirttemberg, and Department of Transfusion Medicine, University of Ulm.

Bernhard Kubanek, Prof Dr Med, Head, Department of Transfusion Medicine, University of Ulm; and Medical Director, German Red Cross Blood Transfusion Service of Baden- Wiirttemberg.

Klaus Koerner, Dr Red Nat, Head of the Infectious Disease

Volume 38, October 1998 TRANSFUSION 907