Commentary

C2.2.1
The satisfactory application of nucleic acid detection techniques depends on the correct performance of all components of testing procedures, including patient preparation and consent (where appropriate), specimen collection, transportation, reagent preparation, nucleic acid isolation, amplification, product visualisation, data transcription, data interpretation, reporting, record keeping, sample storage and quality assurance.

Proficiency and competence

Standard

S2.2.1
The laboratory shall maintain proficiency and competence by undertaking sufficient testing to maintain the knowledge, experience and expertise of staff.

Commentary

C2.2.2
Achieving and maintaining standards of practice and expertise requires an adequate number of samples to be tested with a given method (McGovern et al 1999).

Guideline

G2.2.1
The benefits of centralisation versus those of developing local expertise and autonomy should be carefully assessed by laboratories intending to establish a new nucleic acid detection test.
G2.2.2
The performance of only a small number of tests with any given nucleic acid based method is discouraged.
G2.2.3
Where small-volume testing is necessary, appropriate training, assessment and ongoing education procedures should be implemented to maintain and document the proficiency of staff in conducting these tests, and the number of quality assurance samples should be increased.

Sample collection

Standard

S2.2.2
Trained personnel shall be available at the facility where the sample collection is performed. Specimens shall be collected in accordance with written specimen collection protocols.
S2.2.3
Where patient-collected samples are used for diagnosis, clear instruction shall be provided to the patient to reduce the likelihood of sample contamination.

Commentary

C2.2.3
To minimise the risk of contamination, nucleic acid amplification techniques have some special sample collection and preparation needs, in addition to the usual requirements for pathology testing.
C2.2.4
The precise method of sample collection, initial processing and transportation depends on the specimen concerned and the nucleic acid target (DNA or RNA). Specimens are to be collected according to principles outlined below, as sample contamination can occur at any stage of specimen collection and processing.
C2.2.5
Samples that have been used for other tests before nucleic acid detection testing are at increased risk of contamination. This is a particular risk where the previous test performed was processed with other samples containing the nucleic acid of interest. For example, if samples that have been tested for hepatitis C antibody are cross- contaminated by positive samples, they will yield a false positive result for hepatitis C RNA.

Guideline

G2.2.4
Wherever possible, nucleic acid detection tests should be performed on dedicated samples or on aliquots taken before other tests are performed. Where it is necessary to perform nucleic acid detection tests on samples that have already been used for other purposes and there is a significant risk of cross-contamination, the report should be annotated accordingly and results confirmed on a dedicated sample
G2.2.5
If samples are referred to another laboratory for testing, it is the responsibility of the referring laboratory to ensure that the sample conditions outlined above have been met, and to inform that receiving laboratory if they have not been met.
G2.2.6
Single-use disposable collection equipment should be used.

Sample transport

Standard

S2.2.4
Where there are specific requirements for the transport and handling of samples for nucleic acid detection, these must be documented and available to referring practitioners and laboratories.

Commentary

C2.2.6
National guidelines for the appropriate transport of patient samples are contained in the NPAAC publication Information on the Transport of Pathology Specimens.

Sample preparation

Standard

S2.2.5
Where nucleic acid extraction is required, nucleic acids should be extracted and purified using standard methods.
S2.2.6
The procedures used for nucleic acid isolation from the full range of sample types, collection methods (eg patient versus staff collection) and the condition of specimens received by the laboratory must be validated and procedures detailed in the laboratory methods manual.

Commentary

C2.2.7
The quality of nucleic acid prepared from a specimen has a major effect on the subsequent probability of successfully performing the test.

Guideline

G2.2.7
Where possible controls should be included which assess the adequacy of sample collection and extraction.

Sample integrity

Commentary

C2.2.8
Care needs to be taken to ensure that DNA and RNA remain intact during sample storage, transport and preparation. If the number of target sequences in the sample is very small, a false negative may be obtained if degradation occurs. If the starting material for amplification is mRNA, the sample should be processed as rapidly as possible after collection to minimise RNA degradation by ribonucleases.

Guideline

G2.2.8
Specific instructions for handling samples to minimise nucleic acid degradation should be included in all relevant manuals and should be available to staff in collection centres.
G2.2.9
Where samples of marginal quality, quantity or integrity have been received, the laboratory should notify the referring clinician and seek recollection. Where only a single sample is available, the test is essential and recollection is not possible, the report should be annotated accordingly.
G2.2.10
Class II biological safety cabinets should be used for specimen preparation.

Methods

Commentary

C2.2.9
Laboratory directors are responsible for ensuring the analytic validity of tests before they make them available for use in clinical practice. This is particularly important for nucleic acid detection techniques because of the range of tests, the high sensitivity of the methodology and the potential for variable specificity of amplification and hybridisation procedures.
C2.2.10
The term ‘method’ includes kits, individual reagents, instruments, platforms and software. Elements of methods endorsed ‘research use only’ or ‘not for diagnostic use’ must be validated by the laboratory before use for diagnostic purposes, as outlined in the NPAAC publication Requirements for the Validation of In-House In-Vitro Diagnostic Devices (IVDs).

Test systems

Standard

S2.2.7
Tests shall be validated in accordance with the NPAAC publication Requirements for the Validation of In-House In-Vitro Diagnostic Devices (IVDs).
S2.2.8
Laboratories that have modified kit components or the manufacturer’s procedures must demonstrate equivalence or superiority of the modified procedure before putting the process into routine use. The modified procedure must be treated as an inhouse test for validation purposes.

Commentary

C2.2.11
Commercial kits have been developed for a number of nucleic acid detection tests, particularly for infectious agents. A large number of nucleic acid detection techniques that are currently in use in Australia have been developed within individual laboratories, or adapted from published methods by laboratories.

Guideline

G2.2.11
Where a laboratory uses a commercial test kit in testing in which the methodology and reagents are unchanged from the manufacturer’s instructions, the kit does not need to be independently revalidated in the testing laboratory.
G2.2.12
The supply of some reagent kits is regulated by the Therapeutic Goods Administration under the Therapeutic Goods Act 1989. All kits used for human immunodeficiency virus (HIV) and hepatitis C testing, including nucleic acid detection tests for these agents, are registered on the Australian Register of Therapeutic Goods. Laboratories carrying these tests should ensure that they meet the requirements of these regulations. The range of registered tests and the requirements may change from time to time.
G2.2.13
Both commercial kits and inhouse test systems should be regularly reviewed to ensure that the currency of results is maintained, taking into account new discoveries relevant to the field.

Validation of methods

Standard

S2.2.9
Inhouse tests, or commercial tests endorsed by the manufacturer as ‘for research use only’ or ‘not for diagnostic use’, shall be validated in accordance with the NPAAC publication Requirements for the Validation of In-House In-Vitro Diagnostic Devices (IVDs).
S2.2.10
If any modification is made to a commercial IVD, it must be treated as an inhouse IVD and it shall be fully validated in accordance with the NPAAC publication Requirements for the Validation of In-House In-Vitro Diagnostic Devices (IVDs).
S2.2.11
Validation data shall be retained by the laboratory in sufficient detail to enable external review. The period for which laboratory records are to be retained is stipulated in the NPAAC publication Guidelines for the Retention of Laboratory Records and Diagnostic Materials.

Commentary

C2.2.12
Laboratories should only offer nucleic acid detection tests as a routine test if their technical validity has been established either by the laboratory or the manufacturer.

Guideline

G2.2.14
Where a validated test is available, it should be used in preference to a nonvalidated test. If a nonvalidated test must be performed because of clinical necessity, the report should clearly indicate that the diagnostic validity of the test has not been established.
G2.2.15
The procedures and methods used in nucleic acid detection techniques for diagnostic purposes should be validated for clinical use according to the following principles:
a) Evaluation should use known positive and negative samples.
b) Evaluation should be by comparison with proficiency test material, if available.
c) Evaluation should be by comparison with an existing validated method in the laboratory.
d) Evaluation should be with all specimen types and conditions that will be used to make laboratory diagnosis.
e) If a significant modification to an analytical procedure has been made, the modified procedures should be compared to the original using either an identical sample or identical types of samples.
f ) Reproducibility should not be determined by repetitive analysis of the same sample.
g) The sensitivity of a test, or cutoff values, should be set at a level that is relevant to the diagnostic use of the test.

Contamination control

Commentary

C2.2.13
Nucleic acid detection techniques are usually designed to maximise sensitivity, and are capable of detecting very small amounts of nucleic acid.
C2.2.14
Contamination may occur:
a) during specimen collection or transport
b) during handling or testing in the testing or referring laboratory before nucleic acid detection
c) during extraction of nucleic acids from the sample d) during amplification
e) during product detection
f ) by contamination from reagents used for the test.
C2.2.15
The sources of potential contamination include:
a) positive samples (cross-contamination)
b) amplified nucleic acid (eg contamination of stock reagents or equipment, or in aerosol droplets)
c) operator-derived nucleic acid.
C2.2.16
Mathematical methods are available to assess the likelihood of contamination in the use of PCR-based methods (Shapiro 1999).

Measures to control contamination

Standard

S2.2.12
Laboratories shall retain records documenting contamination events, the identified source of the contamination and measures taken to reduce the risk of future similar contamination events.

Commentary

C2.2.17
The greatest protection a laboratory has against contamination arises from:
a) the competence of the staff in performing laboratory tasks
b) the design of the laboratory
c) the routine use of controls to detect contamination.
C2.2.18
These issues are addressed in detail in the relevant sections of this document.
C2.2.19
For single-round PCRs, the contamination risk may be reduced by replacing thymidine with uracil. The amplified product can then be destroyed by uracil-N-glycosolase, which is added to new samples. This is not sufficient to deal with heavy contamination and is not a substitute for the other measures. Also, it cannot be used in nested PCR, which poses additional contamination risks due to the large amounts of second-round product, which is smaller and more resistant to decontamination procedures. Probe amplification methods (eg branched-chain DNA) have low contamination potential and may be performed in routine laboratory areas, provided those areas are not used for specimen processing or for culture of target microorganisms.

Guideline

G2.2.16
Swabbing of work surfaces and including the swab in the assay can assist in assessing and investigating contamination problems.

Laboratory hygiene

Standard

S2.2.13
Spills shall be cleaned up promptly. Those containing amplified nucleic acids shall be covered with absorbent paper soaked in 2–10% sodium hypochlorite and left for 10 minutes. The absorbent paper shall then be discarded and the area wiped over with 2–10% sodium hypochlorite.
S2.2.14
The level of laboratory hygiene shall be high. Staff shall be careful to avoid contamination of their gowns and gloves. These shall be changed promptly if they have potentially been contaminated, or if they become soiled.

Commentary

C2.2.20
High standards of laboratory hygiene are required, including the measures outlined below. All work areas, especially those used previously for other testing, should be cleaned thoroughly.

Guideline

G2.2.17
Gowns and gloves are to be worn by laboratory staff in the work areas. Gloves are to be discarded and hands washed before leaving the area. Gowns should be dedicated to each area. Gowns or gloves should be changed whenever there is evidence of soiling.
G2.2.18
Tube or pack racks used for holding tubes or plates containing amplified nucleic acid should be decontaminated in 2–10% sodium hypochlorite (or similar reagent) for a minimum of four hours. For screening tests for transfusion or transplantation purposes, decontamination should be according to manufacturer’s recommendations as a minimum or according to any licensing conditions imposed by the Therapeutic Goods Administration.

Equipment

Standard

S2.2.15
All equipment shall be maintained in working order as part of a regular schedule of maintenance in line with the manufacturer’s recommendations.
S2.2.16
Centrifuges used for preparation of biological samples shall have either sealed buckets or a sealed rotor.

Controls

Standard

S2.2.17
At least one of the negative controls and a weak positive control (wherever available) shall be subject to the whole test process, including the extraction.

Commentary

C2.2.21
The types of controls used in nucleic acid detection techniques will vary with specific assays and according to whether microbiological or human molecular genetic testing is being undertaken.
C2.2.22
The exact number of controls required for PCR-based systems depends on the number of samples in each run. A negative patient sample is essential, and laboratories may also include a ‘no nucleic acid’ sample (ie all reagents but no human or organism nucleic acid) to determine whether reagent
contamination is the cause of reactions in the negative patient sample, should they occur. A negative control should be placed after the last patient sample.

C2.2.23
Each run must include a positive control. Ideally, this should be a positive patient sample that has been diluted to a level that is just above the limits of detection for the test and is reliably positive when the test is performing adequately. Usually this control contains target that is at a concentration ten-fold higher than the limit of detection, ensuring that the result will be reliably positive if the assay is performing correctly, but will also detect a loss of sensitivity. If such samples are not available,
a negative patient sample that has had target sequence added — either in the form of cultured organism or synthetic target — may be used. If these cannot be accessed and the test is clinically necessary, a weak extraction control should be added and the report should be annotated to indicate that the sensitivity of the assay could not be confirmed.

C2.2.24
Controls for inhibition are now used routinely in many laboratories in order to reduce the chance of false negatives due to the presence of inhibitory substances in the specimen. These may either be a human gene target that is expected to be present in the sample, a modified target sequence (eg a plasmid with a modified wild-type target sequence) that is added to the sample, or some other target (eg equine herpes virus) that is added to samples and will not yield false positive results. Choosing the appropriate control will vary with specimen type and the ease and availability of control material. For example, using a human gene target may also assess the adequacy of the sample. However some samples, such as cerebrospinal fluid, may not contain enough of that target. Also samples of blood, body fluids or tissues do not require assessment of specimen adequacy. Inhibition controls that are modifications of the wild-type target are theoretically the best but they may not be available. If used, they may compete for primers and reagents and lead to a reduction in sensitivity. Competition for reagents may also occur with the other types of inhibition controls if the inhibition control target is in marked excess compared to the diagnostic target. Laboratories need to consider the merits of the various forms of inhibition control and choose the most appropriate for their test.

Guideline

G2.2.19
Where test runs are expected to contain a large proportion of positive results, it is recommended that additional ‘no template’ controls be interspersed among the patient samples at an appropriate frequency as validated by the particular laboratory.
G2.2.20
Positive controls should be selected to verify that the test would detect significant variants.
G2.2.21
For microbial screening tests using a commercial manufacturer’s reagents, additional controls should be sourced independently of the manufacturer, with levels of target sequence close to the sensitivity of the test but above 95% detection confidence intervals. These should be used in every run.
G2.2.22
Positive controls should be just above the limit of sensitivity of the test.
G2.2.23
Positive controls should be selected to verify that the test would detect significant known genetic variants.
G2.2.24
Inhibition controls should be used wherever possible, especially if particular test and specimen combinations are known to regularly experience problems with inhibition.

Additional testing

Commentary

C2.2.25
Laboratories performing nucleic acid amplification testing should be aware that erroneous results may occur for reasons other than contamination; for example, due to:
a) nonspecific primer binding and amplification of other sequences that are then misidentified as the target sequence
b) amplification of similar or identical target sequences found in other organisms
c) nonspecific, usually low-level, reactivity in the detection system
e) primer design and binding
f ) limitations of the methodologies of detection
g) presence of deletions.
C2.2.26
The frequency and significance of these events will vary with the test
technique, the target sequence chosen, the detection system used and the patient population.

C2.2.27
The use of supplemental testing rests on the positive predictive value of the test and the significance of the result. For those tests with a high positive predictive value, supplemental testing may not be indicated for all samples. For those tests with lower positive predictive value (eg screening tests) supplemental testing is required. Supplemental testing may take the form of further laboratory tests performed on the original sample (eg microscopy, biochemical testing), on the PCR product itself (eg nested PCR, DNA sequencing, diagnostic restriction enzyme digestion) or by further clinical examination and investigation.

Guideline

G2.2.25
It is strongly recommended that where the implications of a positive result are substantial (eg sexually transmitted diseases) and where the specificity of the test is known to be suboptimal or has not been established by extensive validation, one or more of the following methods be used to verify positive results:
a) nested PCR or other techniques that require the binding of more than one set of primers to generate a positive result
b) repeat testing of all positives using another set of primers directed at a different target sequence
c) identification of the product using methods such as specific probes, sequencing or restriction enzyme analysis.
G2.2.26
It may be necessary to use several of these methods to achieve the desired specificity.

Nucleic acid sequencing

Standard

S2.2.18
Laboratories performing sequence-based identification of microorganisms shall have staff who have received specific training in DNA sequencing, sequence editing and database interpretation.

Commentary

C2.2.28
Because nucleic acid sequencing is a rapidly developing area with very few standardised methods, its use requires an additional level of expertise to that needed for other nucleic acid detection methods. This is particularly so in regard to knowledge and expertise in sequencing methods, the editing of sequences, the use of databases for organism identification from sequences and the use of phylogenetic software.

Current databases are largely voluntary and the reliability of organism identification is variable, although the linking of submitted sequences to refereed publications is improving. However, specialised databases using sequences from well-characterised organisms are becoming more common. Interpretation of sequence data also requires knowledge of the natural mutation rates over time and the degree of sequence variability within the target population.
C2.2.29
As sequencing includes a number of additional procedures and transfers of data, laboratories need to be fastidious about maintaining the accuracy, traceability, integrity and security of data.
C2.2.30
The target chosen for sequencing may vary depending on the purpose for which sequencing is performed. The target should be found in all strains and, if the sequence is to be used to type the organism, it should also be sufficiently variable to provide the required level of organism and strain differentiation.
C2.2.31
Ideally, the sequencing should be performed on cloned DNA derived from cultured organisms, because such DNA does not require further amplification before sequencing. However, this is usually not possible or is impractical. If product of PCR or reverse transcription PCR is used for sequencing, variability in sequence may arise from incorporation errors by the DNA polymerase or reverse transcriptase, but this is usually minor. This risk can be reduced either by repeating the amplification and sequencing, or by sequencing both strands of amplified genomic DNA.

Guideline

G2.2.27
Where recommendations have been published by reputable organisations (eg recommendations for HIV and measles genotyping) these should be followed.
G2.2.28
The segment chosen for sequencing should be found in all strains, but should also be sufficiently variable to provide the required level of organism and strain differentiation. Sequences that show excessive natural variation, such as those in quasispecies of HIV and hepatitis C, should be avoided.
G2.2.29
Organisms should be identified based on matches with several sequences from different sources, and preference should be given to sequences supplied from known reputable sources.
G2.2.30
Identification should be based on sequencing of both strands, unless identification based on a single strand has been properly validated.
G2.2.31
Where sequencing is done on DNA amplified directly from patient samples, laboratories must ensure that they check for host DNA sequences that may match the organism.
G2.2.32
Identification should use only high-quality sequences.
G2.2.33
The laboratory should record the database used for the sequence identification and the degree of certainty.
G2.2.34
Regular quality checks should be performed using known organisms (See Fredericks and Relman 1996, Turenne et al 2001).