Appendix XIV L. Nucleic Acid Amplification Techniques

Nucleic acid amplification techniques are based on 2 different approaches:

1. amplification of a target nucleic acid sequence using, for example, polymerase chain reaction (PCR), ligase chain reaction (LCR), or isothermal ribonucleic acid (RNA) amplification;

2. amplification of a hybridisation signal using, for example, for deoxyribonucleic acid (DNA), the branched DNA (bDNA) method; in this case signal amplification is achieved without subjecting the nucleic acid to repetitive cycles of amplification.

In this general chapter, the PCR method is described as the reference technique. Alternative methods may be used, if they comply with the quality requirements described below.

This section establishes the requirements for sample preparation, in vitro amplification of DNA sequences and detection of the specific PCR product. With the aid of PCR, defined DNA sequences can be detected. RNA sequences can also be detected following reverse transcription of the RNA to complementary DNA (cDNA) and subsequent amplification.

PCR is a procedure that allows specific in vitro amplification of segments of DNA or of RNA after reverse transcription into cDNA.

Following denaturation of double-stranded DNA into single-stranded DNA, 2 synthetic oligonucleotide primers of opposite polarity anneal to their respective complementary sequences in the DNA to be amplified. The short double-stranded regions that form as a result of specific base pairing between the primers and the complementary DNA sequence border the DNA segment to be amplified, and serve as starting positions for in vitro DNA synthesis by means of a heat-stable DNA polymerase.

Amplification of the DNA occurs in cycles consisting of:

• — heat denaturation of the nucleic acid (target sequence) into 2 single strands;

• — specific annealing of the primers to the target sequence under suitable reaction conditions;

• — extension of the primers, which are bound to both single strands, by DNA polymerase at a suitable temperature (DNA synthesis).

Repeated cycles of heat denaturation, primer annealing and DNA synthesis results in an exponential amplification of the DNA segment limited by the primers.

The specific PCR product known as an amplicon can be detected by a variety of methods of appropriate specificity and sensitivity.

Multiplex PCR assays use several primer pairs designed for simultaneous amplification of different targets in one reaction.

Because of the high sensitivity of PCR, the samples must be protected against external contamination with target sequences. Sampling, storage and transport of the test material are performed under conditions that minimise degradation of the target sequence. In the case of RNA target sequences, special precautions are necessary since RNA is highly sensitive to degradation by ribonucleases. Care must be taken since some added reagents, such as anticoagulants or preservatives, may interfere with the test procedure.

The risk of contamination requires a strict segregation of the areas depending on the material handled and the technology used. Points to consider include movement of personnel, gowning, material flow and air supply and decontamination procedures.

The system should be sub-divided into compartments such as:

• — master-mix area (area where exclusively template-free material is handled, e.g. primers, buffers, etc.);

• — pre-PCR (area where reagents, samples and controls are handled);

• — PCR amplification (amplified material is handled in a closed system);

• — post-PCR detection (the only area where the amplified material is handled in an open system).

When preparing samples, the target sequence to be amplified needs to be efficiently extracted or liberated from the test material in a reproducible manner and in such a way that amplification under the selected reaction conditions is possible. A variety of physico-chemical extraction procedures and/or enrichment procedures may be employed.

Additives present in test material may interfere with PCR. The procedures described under 7.3.2. must be used as a control for the presence of inhibitors originating from the test material.

In the case of RNA-templates, care must be taken to avoid ribonuclease activity.

PCR amplification of the target sequence is conducted under defined cycling conditions (temperature profile for denaturation of double-stranded DNA, annealing and extension of primers; incubation times at selected temperatures; ramp rates). These depend on various parameters such as:

• — the length and base composition of primer and target sequences;

• — the type of DNA polymerase, buffer composition and reaction volume used for the amplification;

• — the type of thermocycler used and the thermal conductivity rate between the apparatus, reaction tube and reaction fluid.

The amplicon generated by PCR may be identified by size, sequence, chemical modification or a combination of these parameters. Detection and characterisation by size may be achieved by gel electrophoresis (using agarose or polyacrylamide slab gels or capillary electrophoresis) or column chromatography (for example, liquid chromatography). Detection and characterisation by sequence composition may be achieved by the specific hybridisation of probes having a sequence complementary to the target sequence or by cleavage of the amplified material reflecting target-specific restriction-enzyme sites. Detection and characterisation by chemical modification may be achieved by, for example, incorporation of a fluorophore into the amplicons and subsequent detection of fluorescence following excitation.

Detection of amplicons may also be achieved by using probes labelled to permit a subsequent radioisotopic or immuno-enzyme-coupled detection.

A valid result is obtained within a test only if the positive control(s) is unambiguously positive and the negative control(s) is unambiguously negative. Due to the very high sensitivity of the PCR method and the inherent risk of contamination, it is necessary to confirm positive results by repeating the complete test procedure in duplicate, where possible on a new aliquot of the sample. The sample is considered positive if at least one of the repeat tests gives a positive result. As soon as a measurable target threshold is defined, a quantitative test system is required.

The validation programme must include validation of instrumentation and the PCR method employed. Reference should be made to the ICH guidelines (topic Q2B) Validation of Analytical Method: Methodology.

Appropriate official working reference preparations or in-house reference preparations calibrated against International Standards for the target sequences for which the test system will be used are indispensable for validation of a PCR test.

During validation of qualitative tests, the positive cut-off point must be determined. The positive cut-off point is defined as the minimum number of target sequences per volume sample that can be detected in 95 per cent of test runs. The positive cut-off point depends on interrelated factors such as the volume of the sample extracted and the efficacy of the extraction methodology, the transcription of the target RNA into cDNA, the amplification process and the detection.

To define the detection limit of the assay system, reference must be made to the positive cut-off point for each target sequence and the test performance above and below the positive cut-off point.

For a quantitative assay, the following parameters are determined during validation: accuracy, precision, specificity, quantitation limit, linearity, range and robustness.

All reagents crucial for the methodology used have to be controlled prior to use in routine applications. Their acceptance/withdrawal is based on pre-defined quality criteria.

Primers are a crucial component of the PCR assay and as such their design, their purity and the validation of their use in a PCR assay require careful attention. Primers may be modified (for example, by conjugation with a fluorophore or antigen) in order to permit a specific method of detection of the amplicon, provided such modifications do not inhibit accurate and efficient amplification of the target sequence.

In order to minimise the risk of contamination and to ensure adequate sensitivity, the following external controls are included in each PCR assay:

• — positive control: this contains a defined number of target-sequence copies, the number being close to the positive cut-off value, and determined individually for each assay system and indicated as a multiple of the positive cut-off value of the assay system;

• — negative control: a sample of a suitable matrix already proven to be free of the target sequences.

Internal controls are defined nucleic acid sequences containing, unless otherwise prescribed, the primer binding sites. Internal controls must be amplified with defined efficacy, and the amplicons must be clearly discernible. Internal controls must be of the same type of nucleic acid (DNA/RNA) as the material to be tested. The internal control is preferably added to the test material before isolating the nucleic acid and therefore acts as an overall control (extraction, reverse transcription, amplification, detection).

The threshold control for quantitative assays is a test sample with the analyte at a concentration that is defined as the threshold not to be exceeded. It contains the analyte suitably calibrated in International Units and is analysed in parallel in each run of a quantitative assay.

Participation in external quality assessment programmes is an important PCR quality assurance procedure for each laboratory and each operator.

The following sections are published for information.

The majority of nucleic acid amplification analytical procedures are qualitative (quantal) tests for the presence of nucleic acid with some quantitative tests (either in-house or commercial) being available. For the detection of HCV RNA contamination of plasma pools, qualitative tests are adequate and may be considered to be a limit test for the control of impurities as described in the Pharmeuropa Technical Guide for the elaboration of monographs, December 1999, Chapter III 'Validation of analytical procedures'. These guidelines describe methods to validate only qualitative nucleic acid amplification analytical procedures for assessing HCV RNA contamination of plasma pools. Therefore, the 2 characteristics regarded as the most important for validation of the analytical procedure are the specificity and the detection limit. In addition, the robustness of the analytical procedure should be evaluated.

However, this document may also be used as a basis for the validation of nucleic acid amplification in general.

For the purpose of this document, an analytical procedure is defined as the complete procedure from extraction of nucleic acid to detection of the amplified products.

Where commercial kits are used for part or all of the analytical procedure, documented validation points already covered by the kit manufacturer can substitute for the validation by the user. Nevertheless, the performance of the kit with respect to its intended use has to be demonstrated by the user (e.g. detection limit, robustness, cross-contamination).

Specificity is the ability to assess unequivocally nucleic acid in the presence of components that may be expected to be present.

The specificity of nucleic acid amplification analytical procedures is dependent on the choice of primers, the choice of probe (for analysis of the final product) and the stringency of the test conditions (for both the amplification and the detection steps).

When designing primers and probes, the specificity of the primers and probes to detect only HCV RNA should be investigated by comparing the chosen sequences with sequences in published data banks. For HCV, primers (and probes) will normally be chosen from areas of the 5' non-coding region of the HCV genome which are highly conserved for all genotypes.

The amplified product should be unequivocally identified by using one of a number of methods such as amplification with nested primers, restriction enzyme analysis, sequencing, or hybridisation with a specific probe.

In order to validate the specificity of the analytical procedure, at least 100 HCV RNA-negative plasma pools should be tested and shown to be non-reactive. Suitable samples of non-reactive pools are available from the European Directorate for the Quality of Medicines & HealthCare (EDQM).

The ability of the analytical procedure to detect all HCV genotypes will again depend on the choice of primers, probes and method parameters. This ability should be demonstrated using characterised reference panels. However, in view of the difficulty in obtaining samples of some genotypes (e.g. genotype 6), the most prevalent genotypes (e.g. genotypes 1 and 3 in Europe) should be detected at a suitable level.

The detection limit of an individual analytical procedure is the lowest amount of nucleic acid in a sample that can be detected but not necessarily quantitated as an exact value.

The nucleic acid amplification analytical procedure used for the detection of HCV RNA in plasma pools usually yields qualitative results. The number of possible results is limited to 2: either positive or negative. Although the determination of the detection limit is recommended, for practical purposes, a positive cut-off point should be determined for the nucleic acid amplification analytical procedure. The positive cut-off point (as defined in the general chapter 2.6.21 ) is the minimum number of target sequences per volume sample that can be detected in 95 per cent of test runs. This positive cut-off point is influenced by the distribution of viral genomes in the individual samples being tested and by factors such as enzyme efficiency, and can result in different 95 per cent cut-off values for individual analytical test runs.

In order to determine the positive cut-off point, a dilution series of a working reagent or of the hepatitis C virus BRP, which has been calibrated against the WHO HCV International Standard 96/790, should be tested on different days to examine variation between test runs. At least 3 independent dilution series should be tested with a sufficient number of replicates at each dilution to give a total number of 24 test results for each dilution, to enable a statistical analysis of the results.

For example, a laboratory could test 3 dilution series on different days with 8 replicates for each dilution, 4 dilution series on different days with 6 replicates for each dilution, or 6 dilution series on different days with 4 replicates for each dilution. In order to keep the number of dilutions at a manageable level, a preliminary test (using, for example, log dilutions of the plasma pool sample) should be carried out in order to obtain a preliminary value for the positive cut-off point (i.e. the highest dilution giving a positive signal). The range of dilutions can then be chosen around the predetermined preliminary cut-off point (using, for example, a dilution factor of 0.5 log or less and a negative plasma pool for the dilution matrix). The concentration of HCV RNA that can be detected in 95 per cent of test runs can then be calculated using an appropriate statistical evaluation.

These results may also serve to demonstrate the intra-assay variation and the day-to-day variation of the analytical procedure.

The robustness of an analytical procedure is a measure of its capacity to remain unaffected by small but deliberate variations in method parameters and provides an indication of its reliability during normal usage.

The evaluation of robustness should be considered during the development phase. It should show the reliability of the analytical procedure with respect to deliberate variations in method parameters. For NAT, small variations in the method parameters can be crucial. However, the robustness of the method can be demonstrated during its development when small variations in the concentrations of reagents (e.g. MgCl₂, primers or dNTP) are tested. To demonstrate robustness, at least 20 HCV RNA negative plasma pools (selected at random) spiked with HCV RNA to a final concentration of 3 times the previously determined 95 per cent cut-off value should be tested and found positive.

Problems with robustness may also arise with methods that use an initial ultracentrifugation step prior to extraction of the viral RNA. Therefore, to test the robustness of such methods, at least 20 plasma pools containing varying levels of HCV RNA, but lacking HCV-specific antibodies, should be tested and found positive.

Cross-contamination prevention should be demonstrated by the accurate detection of a panel of at least 20 samples consisting of alternate samples of negative plasma pools and negative plasma pools spiked with high concentrations of HCV (at least 10² times the 95 per cent cut-off value or at least 10⁴ IU/mL).

For biological tests such as NAT, specific problems may arise that influence both the validation and the interpretation of results. The test procedures must be described precisely in the form of standard operating procedures (SOPs). These should cover:

• — the mode of sampling (type of container, etc.);

• — the preparation of mini-pools (where appropriate);

• — the conditions of storage before analysis;

• — the exact description of the test conditions, including precautions taken to prevent cross-contamination or destruction of the viral RNA, reagents and reference preparations used;

• — the exact description of the apparatus used;

• — the detailed formulae for calculation of results, including statistical evaluation.

The use of a suitable run control (for example, an appropriate dilution of hepatitis C virus BRP or plasma spiked with an HCV sample calibrated against the WHO HCV International Standard 96/790) can be considered a satisfactory system-suitability check and ensures that the reliability of the analytical procedure is maintained whenever used.

Technical qualification An appropriate installation and operation qualification programme should be implemented for each critical piece of the equipment used. For confirmation of analytical procedure performance after a change of critical equipment (e.g. thermocyclers), the change should be documented by conducting a parallel test on 8 samples of a plasma pool that is spiked with HCV RNA to a final concentration of 3 times the previously determined 95 per cent cut-off value. All results should be positive.

Operator qualification An appropriate qualification programme should be implemented for each operator involved in the testing. To confirm successful training each operator should test at least 8 replicate samples of a plasma pool spiked with HCV RNA to a final concentration of 3 times the previously determined 95 per cent cut-off value. This test (8 replicate samples) should be repeated twice on 2 separate days, i.e. a total of 24 tests performed on 3 different days. All results should be positive.

The European Pharmacopoeia requires that plasma pools used for manufacture of certain products are tested for the presence of B19 virus (B19V) DNA with a threshold concentration that must not be exceeded. In order to comply with these requirements, quantitative NAT tests are preferred. The characteristics regarded as the most important for validation of the quantitative NAT procedure are accuracy, precision, specificity, quantitation limit, linearity and range. In addition, the robustness of the analytical procedure should be evaluated.

This guideline describes methods to validate NAT analytical procedures for assessing B19V DNA contamination of plasma pools based on the ICH guidelines. However, this document may also be used as a basis for the validation of quantitative NAT in general.

For the purpose of this document, an analytical procedure is defined as the complete procedure from extraction of nucleic acid to detection of the amplified products.

Where commercial kits are used for part or all of the analytical procedure, documented validation points already covered by the kit manufacturer can substitute for the validation by the user. Nevertheless, the performance of the kit with respect to its intended use has to be demonstrated by the user (e.g. precision, accuracy, range, robustness).

Accuracy expresses the closeness of agreement between the value that is accepted as either a conventional true value or an accepted reference value and the value found. The accuracy of an assay is dependent on the calibration of the assay and on the variance of the different assay steps. Though it is recommended to establish the accuracy across the specified range of the analytical procedure, the most important assessment of accuracy is in the range of the threshold concentration. In the case of B19V NAT assays for investigation of plasma pools it is recommended to assess the accuracy of the calibrated assay by assaying at least 5 concentrations (dilution factor of 0.5 log) of B19 virus DNA for NAT testing BRP or another material, suitably calibrated in International Units against the actual WHO B19 DNA International Standard, covering the range of the currently recommended threshold concentration of 10.0 IU/µL B19V DNA (e.g. 10⁵ IU/mL, 10^4.5 IU/mL, 10⁴ IU/mL, 10^3.5 IU/mL and 10³ IU/mL), with at least 3 replicates for each dilution. Accuracy should be reported for the different concentrations in terms of percentage determined compared with the known amount of B19V DNA. It should reflect the level of technology of the respective assays, which should also be defined, for example in collaborative studies.

Precision expresses the closeness of agreement between a series of measurements, obtained from multiple sampling of the same homogenous sample. The precision is defined at 3 levels:

• — repeatability expresses the precision under the same operating conditions over a short interval of time (intra-assay precision); it is assessed by using 1 assay and testing 3 replicates of appropriate dilutions of a B19V DNA-positive sample suitably calibrated in International Units and covering the whole quantitative range of the assay; the coefficient of variation for the individual samples is calculated (intra-assay variability);

• — intermediate precision expresses the intra-laboratory variations (inter-assay precision); it is established by assaying replicates (as routinely used for the assay) of appropriate dilutions of a B19V DNA-positive sample suitably calibrated in International Units covering the whole quantitative range of the assay under different circumstances (e.g. different days, different analysts, different equipment, different reagents); the coefficient of variation for the individual samples is calculated;

• — reproducibility expresses the precision between different laboratories (inter-laboratory precision); it is assessed by participation in quantitative collaborative studies on B19V DNA-NAT assays, e.g. under the Proficiency Testing Scheme (PTS), including the comparative analysis of the obtained quantitative results, where appropriate.

Specificity expresses the ability to assess unequivocally nucleic acid in the presence of components that may be expected to be present. The specificity of NAT analytical procedures is dependent on the choice of primers, the choice of probe (for analysis of the final product) and the stringency of the test conditions (for both the amplification and the detection steps).

When designing primers and probes, the specificity of the primers and probes to detect only human B19V DNA should be investigated by comparing the chosen sequences with sequences in published data banks. There should be no major homology found with sequences unrelated to B19V.

The amplified product should be unequivocally identified by using one of a number of methods such as amplification with nested primers, restriction enzyme analysis, sequencing, or hybridisation with a specific probe.

In order to examine the specificity of the analytical procedure, at least 20 B19V DNA-negative plasma pools should be tested and shown to be non-reactive.

Parvovirus B19 genotypes The International Committee on Taxonomy of Viruses (ICTV) has classified representatives of the 3 genotypes as strains of human parvovirus B19. Genotype 1 represents prototype B19V, genotype 2 represents viral sequences like A6, and genotype 3 represents V9-like sequences. By performing sequence alignment with respective B19V genotype sequences available from nucleic acid sequence databases, primers and probes should be designed to detect and quantify consistently the different human parvovirus B19 genotypes. Reference materials should be used to check the approach chosen. Since biological reference preparations reflecting some genotypes might be difficult to obtain, respective plasmid preparations or synthetic nucleic acids may also serve as a characterised target sequence source. However, those cannot be used to validate the extraction procedure.

The quantitation limit is the lowest amount of nucleic acid in a sample that can be determined quantitatively with suitable precision and accuracy. The quantitation limit of the B19V NAT assay is assessed during the repeatability and intermediate-precision studies by limiting dilution analysis. The lowest concentration of target nucleic acids that is quantitated with suitable precision and accuracy is defined.

The linearity of an assay is its ability to obtain test results that are directly proportional to the concentration of the nucleic acid. The linearity of the B19V NAT assay is assessed during the repeatability and intermediate-precision studies by testing replicates of diluted samples with the concentrations covering the whole quantitative range. The interval between the upper and the lower concentration of the target nucleic acid where test results are directly proportional to the concentrations is defined.

The range of an assay is the interval between the upper and the lower concentration of nucleic acid in the sample for which it has been demonstrated that the procedure has a suitable level of precision, accuracy and linearity. The range of the B19V NAT assay is assessed during the repeatability and intermediate-precision studies by testing replicates of diluted samples. The interval between the upper and the lower concentration that can be expressed with an acceptable degree of accuracy and precision is defined.

The robustness of an analytical procedure is a measure of its capacity to remain unaffected by small but deliberate variations in method parameters and provides an indication of its reliability during normal usage. The evaluation of robustness should be considered during the development phase. It should show the reliability of the analytical procedure with respect to deliberate variations in method parameters. For NAT, small variations in the method parameters can be crucial. Nonetheless, the robustness of NAT can be demonstrated during the development of the method when small variations in the concentrations of reagents, for example MgCl₂, primers or dNTP, are tested. To demonstrate robustness, at least 20 B19V DNA-negative plasma pool samples spiked with B19V DNA at the threshold concentration should be tested and found to have acceptable quantitative values.

Cross-contamination prevention should be demonstrated by the accurate detection of a panel of at least 20 samples consisting of alternate samples of plasma pools without B19V DNA or with levels below the threshold concentration (10 samples) and plasma pools spiked with high concentrations of B19V DNA (at least 10² times the threshold level, 10 samples).

For biological tests such as NAT, specific problems may arise that may influence both the validation and the interpretation of results. The test procedures must be described precisely in the form of standard operating procedures (SOPs). These should cover:

• — the mode of sampling (type of container, etc.);

• — the preparation of mini-pools by manufacturers (where appropriate);

• — the conditions of storage before analysis;

• — the exact description of the test conditions including precautions taken to prevent cross-contamination or destruction of the viral nucleic acids, reagents and reference preparations used;

• — the exact description of the apparatus used;

• — the detailed formulae for calculation of results, including statistical evaluation.

The inclusion of an appropriate threshold control (for example, plasma spiked with a B19V DNA sample suitably calibrated in International Units, such as B19 virus DNA for NAT testing BRP) is considered to be a satisfactory system-suitability check and ensures that the reliability of the analytical procedure is maintained whenever used.

Technical qualification An appropriate installation and operation qualification programme should be implemented for each critical piece of the equipment used. For confirmation of analytical procedure performance after a change of critical equipment (e.g. thermocyclers), the change should be documented by conducting a parallel test on 8 samples of a plasma pool that is spiked with a concentration of B19V DNA around the threshold concentration. All results should be acceptable and reflect the features of the assay as determined during the validation phase.

Operator qualification An appropriate qualification programme should be implemented for each operator involved in the testing. To confirm successful training, each operator should test, on 3 separate days, at least 8 replicate samples of a plasma pool that is spiked with a concentration of B19V DNA around the threshold concentration (i.e. a total of 24 samples). All results should be acceptable and reflect the features of the assay as determined during the validation phase.