WO1999037798A1

WO1999037798A1 - Methods for detecting false negative results in blood screening assays using a single aliquot elevated baseline as an internal assay control

Info

Publication number: WO1999037798A1
Application number: PCT/US1999/001536
Authority: WO
Original assignee: Scantibodies Laboratory, Inc.
Priority date: 1998-01-27
Filing date: 1999-01-26
Publication date: 1999-07-29

Abstract

The present invention relates to novel methods for detecting false negative results that can occur with conventional specific binding assays used for screening blood or blood derived products. These erroneous results can occur from either non-sample sources (such as operator error, electromechanical error, and assay component production error), or from sample assay interference (such as endogenous proteins or exogenous interfering chemicals). A single aliquot elevated baseline is used as an internal assay control to detect false negatives without requiring the use of separate assays.

Description

METHODS FOR DETECTING FALSE NEGATIVE RESULTS IN BLOOD SCREENING ASSAYS USING A SINGLE ALIQUOT ELEVATED BASELINE AS AN INTERNAL ASSAY CONTROL

TECHNICAL FIELD

The present invention relates to novel methods for detecting false negative results that can occur with conventional specific binding assays used for screening blood or blood derived products. These false negative results can occur from either method errors, such as operator, mechanical, electronic, or environmental errors, or sample assay interference, such as endogenous proteins. A single aliquot elevated baseline is used as an internal assay control to detect false negatives without requiring the use of separate assays.

BACKGROUND ART

The use of specific binding assays for screening blood and blood derived products has helped prevent the spread of infectious agents. However, even these assays are subject to the problem of false negative results arising from endogenous substances, such as patient antibodies or microbial antigens, i.e., a result that indicates a lower level of an analyte than, in fact, is present, a level that is not considered high enough to warrant disqualifying blood or blood derived products from being able to be administered. For example, hepatitis B virus (HBN) is an infectious agent which is screened for in blood or blood derived products. In some cases, the endogenous presence of HBV antibodies circulating in a patient's blood as part of the patient's immune system response to an HBN infection are responsible for the inability of an HBV immunoassay to detect the HBV, causing a false negative result. In essence, the antibodies used in the HBV assay are unable to bind to the HBV epitopes because the patient's endogenous anti-HBV antibodies have already blocked these binding sites, sometimes known as self immune inhibition. Thus, problems with blood or blood derived products produced from such an HBV positive donor would go undetected, leading to a serious potential for transmitting infectious diseases through the supply of tainted blood and blood derived products. Previously, the only way to be sure that a blood or blood derived product does not have this type of a false negative result has been to assay every screened sample twice, once for the infectious HBV analyte, and once again for the presence of competing endogenous HBV antibodies. Double testing imposes a significant economic burden to blood screening.

US 5,064,755 to Lawrence V. Howard, Jr. et alia discloses a two-site confirmatory assay to solve the problem of false positive results, i.e., a result that indicates a higher level of an analyte than, in fact, is present, a level that is considered clinically significant. Two specific types of antibodies were required, a detector binding partner and a confirmatory binding partner. The detector binding partner attaches to a first epitope on the analyte being assayed. The confirmatory binding partner attaches to a second epitope on the analyte, this binding at the second epitope being such that the detector binding partner cannot bind to the first epitope if the confirmatory binding partner has bound to the second epitope. In the operation of the second assay, a confirmatory binding partner is added to the sample along with the detector binding partner. If a true positive signal occurs, then there is a signal reduction in comparison to an assay run without the use of the confirmatory binding partner. However, a false positive result is indicated if the signal is not reduced. Once again, separate assays must be performed in order to detect this type of false positive result.

In blood screening, false negative results, unlike false positive results, can have dire consequences to public health. The transfusion or infusion of a blood or blood derived product which actually carries clinically significant levels of infectious agents can cause disease transmission. False negatives are particularly troublesome in that blood or blood derived screened products that are identified as being "negative" have the imprimatur of being free of infectious disease agents when in fact they are not. Thus, "unknown" sources of infection do arise.

The mere fact that a product is withdrawn from use as a result of practicing the present invention does not mean that such products are unusable. Further testing of the

- 2 -- withdrawn product is required to verify further the positivity and etiology of that particular product. However, any disadvantage in requiring such testing is far outweighed by the advantages in correcting a far more serious consequence, the passing into the health care system of tainted blood or blood derived products.

DISCLOSURE OF THE INVENTION

The present invention relates to a single aliquot assay method for detecting a false negative result in a specific binding assay used for screening blood or blood derived products based upon the minimum recovered assay value resulting from the addition of a known amount of a specific binding analyte. Any existing conventional specific binding assay used in screening blood or blood derived products can be used according to the present method so as to detect false negatives without necessarily modifying the reagents of these assays. The existing method can be used in either conventional automated multisample analyzers or manual assays, as well as disposable lateral flow i munochromatographic devices or immunoreactive biosensors. Analytes may comprise antibodies, antigens, nucleic acid sequences, nucleic acid probes and the like. Five steps are involved.

First, one mixes a biological fluid or cellular screening sample with a known, non- endogenous amount of the analyte one is seeking to assay or measure, a complementary specific binding partner, and a signal reagent. (For the purposes of this invention, "added analyte" includes analogues of analytes or analyte complexes that can be detected by the assay for the analyte itself.) The sample must be able to have present the analyte for which a blood or blood derived product is being screened, thus blood, serum, or saliva are suitable sample materials. The known amount of added analyte should be sufficient to produce a signal of at least about twice the standard deviation of the mean for the zero value replicates of the assay when performed without the addition of analyte. Moreover, the known amount of analyte should not exceed a level where one would no longer be able to detect a positive result as well as a false negative result. Thus, the combination of known added analyte and minimum sample analyte level for detecting a positive result should not exceed the functional range for the assay. Because different assay manufacturers construct their assays differently, these values will also differ from assay to assay, but can be easily determined by those of ordinary skill in the art.

The contacting or mixing of sample and added analyte should occur under a predetermined set of conditions normally used for running the assay. The added analyte can be added either directly to the sample individually or incorporated in an assay reagent. There is no particular order of addition that is needed. In some instances, such as with biosensors or lateral flow immunochromatographic devices, the added analyte may already be present about the sample addition site or adjacent thereto. Specific binding partner/analyte complexes are formed which can be detected by measuring the amount of a signal resulting from the presence of the complexes.

Second, one measures the signal from the specific binding partner/analyte complexes. The signal measurement is accomplished through conventional means known to those of skill in the art, and will vary depending upon the type of signal and the particular assay format that is employed.

Third, one determines or calculates the elevation of the assay baseline level that should occur to the sample if there was not any detectable endogenous analyte present, but accounts for the added analyte.

Fourth, one compares the measured signal to the elevated baseline level.

Finally, one detects a false negative result if the signal is substantially below the calculated elevated baseline signal. (For the purposes of this invention, "substantially below" means an increment at least greater than the precision limit corresponding to that level of analyte that is detected.) True negative results will fall between the elevated baseline level and the new cutoff value which incorporates the contribution of the added analyte, which can easily be established by the assay manufacturer in the case of

4 - commercially available assays. False negative results will fall below the elevated baseline. One should note that the present method is not dependent on any particular method for calculating the new cutoff and the elevated baseline. This calculation can be based on statistical models that incorporate variations in individuals with particular disease states.

The present invention can detect false negative results due to endogenous antibodies. Also, it can detect such results where antibodies are the target analyte and endogenous antigen causes a false negative reading by binding inactivation. In order for antibody blood screening assays, such as those used commercially for HIV, HCV, and Chlamydia detection, to function properly, a patient's antibodies must maintain their ability to bind to microbial type epitopes which are presented as an assay component. If the patient sample contains antibodies which have lost that binding ability, then those samples will be read as being negative where, in fact, the samples should register as positive, a false negative. The loss of antibody binding ability can arise where the blood level of the microbe is so high that all of the antibody/analyte in the blood is bound, as in the early stages of some infectious diseases. In other words, there is an excess of endogenous microbial epitopes to endogenous antibodies. In some samples, the endogenous antibodies that interfere with detecting a true positive result are heterophilic antibodies or autoimmune antibodies. An autoimmune antibody, such as rheumatoid factor, can bind to an antibody analyte, blocking the reaction between the analyte and the detecting antibody. A heterophilic antibody, such as human anti-mouse antibody, can bind to the signal antibody of an assay, preventing the binding reaction to the analyte. In either case, a false negative result will occur.

The present invention can detect such false negatives because by adding non- endogenous antibodies, (the added analyte), these excess endogenous epitopes will bind up the added analyte. With the added analyte unable to participate in the assay binding reaction, there will not be an elevated baseline level, and thus, the false negative will be detected by falling below the elevated baseline level..

- 5 The present method can also detect false negative results which arise from non- sample sources, such as mechanical error, assay component production error, or operator error. For example, mechanical errors can include a too low incubation temperature or a signal detector malfunction. Assay component production errors can include defective sample containers or non-homogeneous coating of solid phases. Operator errors include non-homogenous mixing of conjugate solution, insufficient sample mixing, or inaccurate dispensing of reagent, such as adding too little signal generating reagent. In such a case, if the conjugate solution was under-pipetted, then the signal generated would be lower than what an accurate test for that amount of endogenous analyte would generate, and could be falsely undervalued. Moreover, the present method also can detect false negative results which arise from sample collection errors. Certain anticoagulants and preservatives can interfere negatively with the function of an assay. For example, sodium azide can block the enzymatic activity of peroxidase based immunoassay signal generation systems, which could cause a false negative.

The present invention can be applied to many conventional specific binding assays. It is not limited to assay formats that use solid phases. It can be incorporated into homogeneous assays in which analyte dependent signal is generated without a solid phase. It can function with competitive assays where only one binding partner is required. It can function with inhibition type assays where signal inhibition is dependent on analyte.

By employing the present invention, blood banks and other commercial entities that sell or distribute blood or blood derived products can eliminate many, if not most, of the infectious products which currently are in the pipeline for administration due to an erroneous negative identification. Moreover, one can do so without incurring the burdensome expense of running additional separate assays for all screened blood or blood derived products. Lastly, the present invention can be easily implemented through existing blood screening assays and does not entail the financial burden of adding new assays to the blood screening menu or assay profile.

- 6 BRIEF DESCRIPTION OF THE DRAWINGS

FIGURE 1A is a diagrammatic view of an immunoassay for HBV using the present invention.

FIGURE IB is a diagrammatic view of an immunoassay for anti-HCV using the present invention.

FIGURE 2 A is a graph of true negative samples using an existing HBV assay method.

FIGURE 2B is a graph of true negative samples using the assay of FIGURE 2 A according to the present method.

FIGURE 3A is a graph of true positive samples using an existing HBV assay method.

FIGURE 3B is a graph of true positive samples using the assay of FIGURE 3 A according to the present method.

FIGURE 4 A is a graph of true positive samples before adulteration with antibodies using an existing HBV assay method.

FIGURE 4B is a graph of the samples of FIGURE 4 A after adulteration using the same existing HBV assay method, showing false negative results.

FIGURE 4C is a graph of the adulterated samples of FIGURE 4B when tested with the existing HBV assay method used according to the present method.

FIGURE 5 is a graph of false negative samples caused by an exogenous factor,

(interfering chemicals), comparing an existing HBV assay method with the same

_ 7 _ assay used according to the present method.

FIGURE 6 A is a graph of true negative samples using an existing anti-hepatitis C virus antigen (anti-HCV) assay method.

FIGURE 6B is a graph of the samples of FIGURE 6 A using the same assay according to the present method.

FIGURE 7 is a graph of true positive samples using an existing anti-HCV assay method according to the present method.

FIGURE 8 is a graph of false negative samples caused by an exogenous factor, (interfering chemicals), using an existing anti-HCV assay method according to the present method.

FIGURE 9 A is a graph of true negative samples using an existing Human Immune Deficiency Virus Types 1 & 2 antibody (anti-HTV) assay method.

FIGURE 9B is a graph of the samples of FIGURE 9 A using the same assay according to the present method.

FIGURE 10 is a graph of true positive samples using an existing anti-HIV assay method according to the present method.

FIGURE 11 is a graph of false negative samples caused by an exogenous factor using an existing anti-HIV assay method according to the present method.

8 -- BEST MODES FOR CARRYING OUT THE INVENTION

The present invention can be used to detect false negative results for infectious viral agents routinely of interest in screening blood or blood derived products. Such agents include adenovirus, cytomegalovirus, hepatitis A virus, HBV, hepatitis C virus, herpes simplex virus 1 and 2, human immune deficiency viruses, parvo virus, and respiratory syncytial virus. Besides infectious viral agents, the present invention can also be used with conventional screening specific binding assays for infectious microbial agents such as Borrellia, Brucella, Chlamydia trachymatis and psittaci, Clostridium difficiles, Helicobacter pylori, Mycoplasma pneumoniae, rubella, syphilis, and Toxoplasma Gondii.

FIGURE 1 A illustrates schematically how one uses the present invention in a conventional coated bead enzyme immunoassay, in this case a sandwich immunoassay where the analyte is an antigen. A known amount of analyte (12) is added to the patient sample along with the assay mixture of anti-analyte antibody (14) attached to a solid phase (16) and a signal element comprised of a second anti-analyte antibody (18) having a label attached (20). Added analyte can be contacted or mixed at any stage in a conventional assay format by means of any medium wherein the added analyte behaves similarly to any endogenous analyte in creating an assay signal. Any of a number of signal generating systems can be used to measure the bound analyte, including, but not limited to, radioisotopes, colorimetric enzyme reactions, chemiluminescent reactions, and fluorescence.

FIGURE IB illustrates schematically how one uses the present invention in a conventional coated bead enzyme immunoassay, in this case a serological immunoassay where the analyte is an antibody. A known amount of analyte antibody (12) is added to the patient sample along with the assay mixture of anti-analyte antigen (14) attached to a solid phase (16) and a signal element comprised of an anti-analyte antibody (18) having a label attached (20). Added analyte can be contacted or mixed at any stage in a conventional assay format by means of any medium wherein the added analyte behaves similarly to any endogenous analyte in creating an assay signal. Any of a number of signal generating systems can be used to measure the bound analyte, including, but not limited to, radioisotopes, colorimetric enzyme reactions, chemiluminescent reactions, and fluorescence.

HEPATITIS B VIRUS SCREENING ASSAY

A conventional, U. S. Food and Drug Administration (FDA) approved blood screening assay, specifically hepatitis B surface antigen (HBsAg) immunoassay, was used to demonstrate the effectiveness of the present invention. The selected test was a widely used, commercially available coated bead, sandwich enzyme immunoassay test for human serum or plasma samples, which is available from Abbott Labs, Abbott Park, Illinois, and is known as "Auszyme® Monoclonal". The cutoff calculation for this test is the negative control mean absorbance, i.e., NCx, the baseline, plus 0.050. Thus, if a sample registers a signal of less than or equal to this cutoff value, then the sample is defined to be non- reactive or negative for HBsAg.

Forty nine samples were divided into two aliquots, one for testing using the manufacturer's instructions, and one for using in accordance with the present method. Twenty of the samples were true hepatitis B surface antigen (HBsAg) negative. In other words, no detectable endogenous HBsAg was present. Seven of the samples were true HBsAg positive. In other words, detectable endogenous HBsAg was present. The remaining twenty two samples were false negatives caused both by endogenous proteins, such as anti-HBsAg- and exogenous sources, such as interfering chemicals.

True Negatives

The results of the HBsAg testing of the twenty true HBsAg negative serum samples obtained from human patients are shown in FIGURES 2 A and 2B. These samples were identified as being negative by testing them with the Auszyme® Monoclonal HBsAg test. In accordance with the manufacturer's instructions, the first aliquots of the true negative samples (which did not have any known amount of HBsAg added) gave signals which

10 corresponded to a negative result, i.e., the signals were all below the NCx plus 0.050, the cutoff value established by the manufacturer for indicating a negative result. (See FIGURE

2A.)

When the second aliquots of these twenty samples had about 0.105 ng of F-BsAg added and were tested in accordance with the manufacturer's instructions, the signals were elevated to between 0.080 and 0.130 OD units, as shown in FIGURE 2B. For such an amount of added HBsAg, the baseline level should be elevated to at least 0.080 OD units, the calculated elevated baseline. Thus, the new cutoff value was 0.130 OD units (0.080 plus 0.050). In other words, with this assay, a true negative result is one in which a sample reads greater than 0.080 OD units, but less than 0.130 OD units, while a true positive reads greater than 0.130 OD units, and a false negative reads less than 0.080 OD units. All of the second aliquot samples were below this new cutoff value and above the calculated elevated baseline level, and thus, the present invention identified all true negatives as such.

True Positives

The results of the testing of the seven true positive HBsAg serum samples obtained from human patients are shown in FIGURES 3 A and 3B. These samples were identified as being positive by testing them with the Auszyme® Monoclonal HBsAg test. In accordance with the manufacturer's instructions, the first aliquots of the true positive samples (which did not have any known amount of HBsAg added) gave signals which corresponded to a positive result, i.e., the signals were all above the NCx plus 0.050, the cutoff value established by the manufacturer for indicating a positive result. (See FIGURE 3A.)

When the second aliquots of these seven true HBsAg positive samples had 0.105 ng of HbsAg added and were tested in accordance with the manufacturer's instructions, the signals were elevated to between 0.150 OD units and 0.200 OD units, as shown in FIGURE 3B. For such an amount of added HBsAg, the baseline level should be elevated to 0.080 OD units, as explained above. All of the second aliquot samples were above this new cutoff value of 0.130 OD units described above, and thus, the present invention

-- 11 - identified all true positives as such.

False Negatives

The results of the testing of six of the twenty two false HBsAg negative, anti-

HBsAg antibody positive serum samples obtained from human patients are shown in FIGURES 4A, 4B, and 4C. The first six false negative samples were derived from true positive samples adulterated with added human anti-HBsAg. Such antibodies normally appear endogenously in the blood following an HBsAg infection. FIGURE 4A shows the HBsAg assay results of these six samples before adulteration. These samples were assayed with the Auszyme® Monoclonal HBsAg test, in accordance with the manufacturer's instructions, and all were found to be positive. The added human anti-HBsAg was obtained from patient samples and were identified by a conventional anti-HBsAg assay, also made by Abbott Labs. One part of positive human anti-HBsAg sample was added to one part of each HBsAg positive sample to serve as "endogenous" antibodies. FIGURE 4B shows the results from the testing of the samples after adulteration. All of the formerly HBsAg positive samples now register as HBsAg negative, i.e., are false negative results.

When the second aliquots of these six samples had about 0.105 ng of HbsAg added and were tested in accordance with the manufacturer's instructions, the signals for the conventionally "negative" result samples were not elevated to at least 0.080 OD units. (See FIGURE 4C.) For such an amount of added HBsAg, the baseline level should be elevated to at least 0.080 OD units, for a new cutoff value of 0.13 OD units (0.080 plus 0.050). Thus, these false negative samples can be identified as such by the failure of the signals to be elevated in accordance with the known amount of HBsAg added to the mixture. The results from these samples were substantially below the elevated baseline level in that the standard deviation for this assay as used at this level was about 9% as obtained by assaying replicates of a sample containing analyte generating a signal corresponding to the approximate level of the elevated baseline. The measured results fell substantially below the elevated baseline level. These samples were tested for endogenous anti-HBsAg antibodies, and indeed were found to be positive for these antibodies. Thus, the present

12 method can detect a sample containing infectious antigen which is masked by the presence of endogenous antibodies in the sample, most likely arising from a patient's immune response to an infectious antigen.

The results of the testing of the remaining sixteen of the twenty two false negative

HBsAg serum samples obtained from human patients is shown in FIGURE 5. These sixteen samples were prepared to have exogenous factors present in that originally all of these samples were positive for HBsAg as determined by a conventional assay, but each sample had 10 mg of sodium azide added into 0.1 ml of the original sample. Sodium azide inhibits the function of the peroxidase label in the assay. According to the manufacturer's instruction, if sodium azide is present, then a two step assay is needed as opposed to a one step assay. The additional washing step removes sodium azide from the presence of the assay conjugate which includes the peroxidase. If a sample was mislabeled or misread as not containing sodium azide, then the assay could be inadvertently run without using the required washing step. The first aliquots of these sixteen samples also gave negative results when assayed with the Auszyme® Monoclonal HBsAg test, according to conventional instructions, even though endogenous HBsAg was present. In other words, the signals by the conventional assay method were all below the cutoff value of NCx plus 0.050.

When the second aliquots of these sixteen samples had 0.105 ng of FtbsAg added and were tested in accordance with the manufacturer's instructions, the signals for the conventionally "negative" result samples were not elevated to at least 0.080 OD units, as described above and as shown in FIGURE 2B. For such an amount of added HBsAg, the baseline level should be elevated to at least 0.08 OD units. The new cutoff value should be about 0.130 OD units (0.080 plus 0.050). Thus, these samples which are identified as "negative" by conventional assay can be identified as actually being false negatives by the present method due to the failure of the signals to be elevated to at least 0.080 OD units, in accordance with the known amount of HBsAg added to the mixture.

- 13 - HEPATITIS C VIRUS ANTIBODY SCREENING ASSAY

A widely used, conventional FDA approved anti-HCV immunoassay was used as well to demonstrate the effectiveness of the present invention. The selected test was a commercially available recombinant HCV antigen coated bead, sandwich enzyme immunoassay test for human serum or plasma samples, which is available from Abbott Labs, Abbott Park, Illinois, and is known as "Abbott HCV EIA 2.0". The cutoff calculation for this test is the negative control mean absorbance, i.e., NCx, the baseline, plus 0.25 times the mean absorbance of the positive control (PCx). At an OD 492 nm, the NCx value was 0.146, the PCx was 1.195, and the cutoff value was 0.445. Thus, if a sample registers a signal of less than or equal to 0.445, then the sample is determined to be negative for anti-HCV antibodies. In commercial use by blood banks, units of blood are released for transfusion and further processing based on this identification.

Twenty nine samples were tested for anti-HCV antibodies. Seventeen true anti-

HCV negative samples were divided into two aliquots, one for testing using the manufacturer's instructions, and one for using in accordance with the present method. An additional six true positive anti-HCV samples were tested in accordance with the present method. The remaining six samples were false negatives caused by the addition of exogenous factors which influenced the test results, namely sodium azide was present, inhibiting the function of the peroxidase label in the assay exogenous proteins.

True Negatives

The results of the anti-HCV testing of the seventeen true anti-HCV negative serum samples obtained from human patients are shown in FIGURES 6 A and 6B. These samples were identified as being negative by testing them with the Abbott HCV EIA 2.0 test. In accordance with the manufacturer's instructions, the first aliquots of the true negative samples (which did not have any known amount of anti-HCV positive serum added) gave signals which corresponded to a negative result, i.e., the signals were all below the NCx plus 0.25 PCx, the cutoff value established by the manufacturer for indicating a negative

14 result. (See FIGURE 6A.)

The second aliquots of these seventeen samples had added about one part of a human patient plasma sample for anti-HCV antibodies added to 330 parts of each of the seventeen samples, an "anti-HCV additive". This dilution was selected as being sufficient to effect the elevated baseline. Then they were tested in accordance with the manufacturer's instructions. The signals were elevated to between about 0.240 OD units and about 0.350 OD units, as shown in FIGURE 6B. For such an amount of added anti- HCV antibodies, the baseline level should be elevated to at least 0.240 OD units, the calculated elevated baseline. Therefore, the new cutoff value should be about 0.540 OD units, (0.240 plus 0.300 (0.25 times PCx )). In other words, with the addition of the anti- HCV additive, a sample that reads less than 0.540 OD units but greater than 0.240 OD units is a true negative, a sample that reads above 0.540 OD units is a true positive, and a sample that reads less than 0.240 OD units is a false negative. All of the second aliquot samples were below this new cutoff value and above the calculated elevated baseline level, and thus, the present invention identified all true negatives as such.

True Positives

The results of the testing of the six true positive anti-HCV serum samples obtained from human patients is shown in FIGURE 7. These samples were identified as being positive by testing them with the Abbott HCV EIA 2.0 test. The true positive samples had the anti-HCV additive added to each. Prior to the addition, all of the samples had readings above 0.445 OD units. After the addition, the values ranged from 0.620 OD units to 1.000 OD units. The signals were all above the elevated baseline level of about 0.240 OD units and above new cutoff value of about 0.540 OD units. Thus, these six samples were true positives.

15 False Negatives

Six anti-HCV positive serum samples obtained from human patients were identified as being positive by using the Abbott HCV EIA 2.0 test in accordance with the manufacturer's instructions. The samples were then adulterated by adding 10 mg of sodium azide into 0.1 ml of each anti-HCV positive sample. The addition of sodium azide would have rendered such a positive sample to read as negative when used according to conventional methods. When tested in accordance with the manufacturer's instructions, the signals now were changed to false negatives. As shown in FIGURE 8, these false negative samples can be identified as such by the failure of the signals to be elevated up to the elevated baseline level. The results from these samples (all less than 0.180 OD units) were substantially below the elevated baseline level (0.2400 OD units) in that the standard deviation for this assay as used at this level was about 5%, as obtained by assaying replicates of a sample approximately generating a signal corresponding to the elevated baseline. A measurement of 0.180 OD units is 0.12 OD units less than the elevated baseline level, a difference which is greater than 5% of 0.240 OD units. The measured results fell substantially below the elevated baseline level, and, therefore, were identified as false negatives.

HUMAN IMMUNE DEFICIENCY VIRUS TYPES 1 & 2 ANTIBODY

SCREENING ASSAY

A routinely employed, conventional FDA approved anti-HIV immunoassay was used as well to demonstrate the effectiveness of the present invention. The selected test was a widely used, commercially available recombinant HIV-1 env, HTV-1 gag, and HIV-2 env protein coated bead, sandwich enzyme immunoassay test for human serum or plasma samples, which is available from Abbott Labs, Abbott Park, Illinois, and is known as "HIVAB® HTV-l/HTV-2 (rDNA) EIA". The cutoff calculation for this test is the negative control mean absorbance, i.e., NCx, the baseline, plus 0.100. At an OD 492 nm, the NCx value obtained was 0.014, and the cutoff value was 0.114. Thus, if a sample registers a signal of less than or equal to 0.114, then the sample is determined to be non-reactive or

- 16 negative for anti-HIV antibodies. In commercial use by blood banks, units of blood are released for transfusion and further processing based on this identification.

Twenty eight samples were tested for anti-HIV antibodies. Eighteen true anti-HIV negative samples were divided into two aliquots, one for testing using the manufacturer's instructions, and one for using in accordance with the present method. An additional six anti-HIV true positive samples were tested in accordance with the present method. The remaining four samples were false negatives caused by the addition of exogenous factors which influenced the test results, namely sodium azide was present, inhibiting the function of the peroxidase label in the assay.

True Negatives

The results of the anti-HIV testing of the eighteen true anti-HIV negative serum samples obtained from human patients is shown in FIGURES 9A and 9B. These samples were identified as being negative by testing them with the HIVAB® HIV-l/fflV-2 (rDNA) EIA test. In accordance with the manufacturer's instructions, the first aliquots of the true negative samples (which did not have any known amount of anti-HIN positive serum added) gave signals which corresponded to a negative result, i.e., the signals were all below the ΝCx plus 0.100, the cutoff value established by the manufacturer for indicating a negative result. (See FIGURE 9A.)

The second aliquots of these eighteen samples had added about one part of a human patient plasma sample which had tested positive for anti-HIV antibodies to 30,000 parts of each sample, the "anti-HIV additive". This dilution was selected as being sufficient to effect the elevated baseline to a desired level. After the addition, the samples were tested in accordance with the manufacturer's instructions. The signals were elevated to between about 0.120 OD units and about 0.170 OD units, as shown in FIGURE 9B. The elevated baseline level was about 0.120 OD units. A new cutoff value was about 0.220 OD units, (0.120 plus 0.100). In other words, for these samples, a reading of greater than 0.120 OD units but less than 0.220 OD units is a true negative, a reading of greater than 0.220 OD

- 17 units is a true positive, and a reading of less than 0.120 OD units is a false negative. All of the second aliquot samples were below this new cutoff value and above the calculated elevated baseline level, and thus, the present invention identified all true negatives as such.

True Positives

The results of the testing of the six true positive anti-HCV serum samples obtained from human patients is shown in FIGURE 10. These samples were identified as being positive by testing them with the HTVAB® HIV-l/HIV-2 (rDNA) EIA test. The true positive samples had added about one part of the anti-HIV additive to about 30,000 parts of sample. After the addition, they were tested in accordance with the manufacturer's instructions for the HIVAB® HIV-l HIV-2 (rDNA) EIA test. The signals were elevated to between about 0.180 OD units and about 0.290 OD units, as shown in FIGURE 10 The signals were all above the new cutoff value of about 0.220 OD units,, and thus, were identified as true positives.

False Negatives

Six anti-FflN positive serum samples obtained from human were tested with the HIVAB® HIV- 1 IMY-2 (rDΝ A) EIA 2.0 test, in accordance with the manufacturer' s instructions and were identified as being positive. These six samples were adulterated by adding 10 mg of sodium azide into 0.1 ml of each sample and the anti-HIV additive. About one part of the anti-HIV additive was added to about 30,000 parts of each sample. The addition of sodium azide would have rendered such a positive sample to read as negative when used according to conventional methods. The samples were tested with the HIVAB® HTV-l/fflV-2 (rDΝA) EIA 2.0 test, in accordance with the manufacturer's instructions. The results were between about 0.040 OD units and about 0.050 OD units, as shown in FIGURE 11. The baseline level for the assay of the adulterated samples was elevated to at least about 0.120 OD units. The new cutoff value was about 0.220 OD units (0.120 plus 0.100). The results from these samples were substantially below the elevated baseline level in that the standard deviation for this assay as used at this level was about

- 18 - 8%, as obtained by assaying replicates of a sample approximately generating a signal corresponding to the desired elevated baseline level. The measured results fell substantially below the elevated baseline level, and thus, were identified as false negatives.

The ordinarily skilled artisan can appreciate that the present invention can incorporate any number of the preferred features described above.

All publications or unpublished patent applications mentioned herein are hereby incorporated by reference thereto.

Other embodiments of the present invention are not presented here which are obvious to those of ordinary skill in the art, now or during the term of any patent issuing from this patent specification, and thus, are within the spirit and scope of the present invention.

- 19 -

Claims

WE CLAIM:

1. A method for detecting a false negative result from a specific binding assay used for screening blood or blood derived products comprising: a) contacting a sample for a blood or blood derived product to be screened with a predetermined and known amount of non-endogenous analyte, a complementary specific binding partner, and a signal reagent under a predetermined set of conditions, whereby specific binding partner/analyte complexes are formed which can correspond to the amount of analyte present and can be detected by measuring the amount of a signal resulting from the presence of the complexes; b) measuring the signal from the specific binding partner/analyte complexes; c) determining an elevated baseline level for the assay that is based on an expected increase in signal due to the added non-endogenous analyte; d) comparing the measured signal to the elevated baseline level; and e) detecting a false negative result for the sample if the measured signal is substantially below the elevated baseline level.

2. The method of Claim 1 wherein the analyte is a virus selected from the group consisting of adenovirus, cytomegalovirus, HBV, hepatitis A virus, hepatitis C virus, herpes simplex virus 1, herpes simplex virus 2, human immune deficiency viruses, parvo virus, and respiratory syncytial virus.

3. The method of Claim 1 wherein the analyte is a bacterium selected from the group consisting of Borrellia, Brucella, Chlamydia trachymatis and psittaci, Clostridium difficiles, Helicobacter pylori, Mycoplasma pneumoniae, rubella, syphillis, and Toxoplasma Gondii.

4. The method of Claim 1 wherein the analyte is a specific binding partner to a microbe.

5. The method of Claim 4 wherein the analyte is an antibody and endogenous epitopes on the microbe bind to substantially all of the antibody epitopes present in the sample which

~ 20 - participate in a specific binding reaction involving the complementary specific binding partner and the signal reagent.

6. The method of Claim 4 wherein the microbe is a virus.

7. The method of Claim 6 wherein the analyte is an antibody and endogenous epitopes on the virus or viral components bind to substantially all of the antibody epitopes present in the sample which participate in a specific binding reaction involving the complementary specific binding partner and the signal reagent.

8. The method of Claim 1 wherein the complementary specific binding partner also comprises the signal reagent.

21 -