Establishment of the 1^st World Health Organization International Standard for Plasmodium falciparum DNA for nucleic acid amplification technique (NAT)-based assays

Each year there are an estimated 300 million acute cases of malaria worldwide, accounting for more than one million deaths annually [1]. In humans, malaria is caused by infection with five species of Plasmodium (Plasmodium falciparum, Plasmodium malariae, Plasmodium ovale, Plasmodium vivax and Plasmodium knowlesi). Plasmodium falciparum is associated with the greatest morbidity and mortality, accounting for approximately 95% of deaths due to malaria [2]. Although most of these cases are found in endemic tropical regions of Africa, Asia, Central and South America, cases of imported malaria are reported in non-endemic areas. In Europe it has been estimated that there are in excess of 10,000 cases of imported malaria per year, with around 2,000 cases being recorded annually in the UK alone [3, 4]. Estimates by the Centers for Disease Control and Prevention (United States) have indicated that there were approximately 1,528 cases, including seven fatal cases, of imported malaria in the United States in 2005, largely due to travellers, immigrants and servicemen returning from malaria endemic regions [5]. Transfusion transmitted cases of malaria can also occur, by the use of contaminated blood, and may result in significant disease or even death in the recipients, particularly when P. falciparum is the causative agent [6].

Malaria diagnosis has relied upon microscopic analysis of Wright's or Wright-Giemsa stained blood smears to detect infection and determine parasite loads [7]. Slide interpretation requires considerable expertise and is difficult to standardize, particularly at low levels of parasitaemia or where mixed species infections occur [8, 9]. Nucleic acid amplification technique (NAT)-based assays, such as PCR are becoming increasingly employed in the diagnosis of malaria [10, 11]. Studies have shown that PCR can improve sensitivity and species discrimination, when compared to either microscopy or serological methods [12]. Although numerous PCR methods have been developed for the laboratory diagnosis and clinical management of malaria, the reported sensitivities of these assays vary widely. These differences may be a result of intrinsic variability in assay sensitivity or a consequence of calibration using different reference reagents, which are poorly standardized. In terms of quality assurance, it is of concern that variability in the implementation of assays between laboratories could give different results for the same patient sample.

In the last ten years, a number of International Standards have been developed by the World Health Organization (WHO) for NAT-based assays. The first was established in 1997 for hepatitis C virus (HCV) RNA [13]. This standard was pivotal in the introduction of HCV RNA NAT screening of blood and plasma donations, used in transfusion and the manufacture of plasma derived medicinal products, where previously there had been transmissions of HCV. Further standards have been established for hepatitis B virus (HBV) DNA, HIV-1 RNA, parvovirus B19 DNA, hepatitis A virus (HAV) RNA [14‐17]. The standards have allowed the development of regulatory requirements for the detection of these blood borne viruses in plasma, by setting commonly recognized thresholds and limits of detection. The standards have been used in the calibration of secondary standards and working reagents, and have been used in the validation of assays for both the qualitative and quantitative NAT-based techniques. These blood-borne virus standards have also found use in the clinical setting, both for diagnosis and for monitoring viral loads in response to antiviral therapy. Moreover, results in a number of commercially available assays for blood borne viruses are expressed in terms of WHO International Units associated with the respective standards.

The use of NAT-based assays for the detection of P. falciparum DNA in blood is now being increasingly used in the management and diagnosis of malaria, particularly in countries where travellers are returning from malaria endemic regions. Such NAT-based assays are also being used, on occasion in the primary screening or confirmatory testing of blood used for transfusion [Dr. S. Salueda, personal communication]. In some centres performing vaccine trials, there are reports of using DNA-based assays to monitor parasitaemia using real-time PCR [26]. However there is widespread variation in the sensitivity and specificity of these assays and the aim of this study was to evaluate candidate reference preparations for P. falciparum DNA that will be useful for standardization purposes.

The candidate standards were evaluated by a wide range of assays from a large number of laboratories worldwide, as is the requirement of such WHO international collaborative studies [27]. During this study, the participating laboratories used a variety of different extraction and amplification methods for the detection of P. falciparum DNA (Table 2). The majority of these assays were developed in-house and therefore it was not unsurprising to see such a spread of results reported. The participating laboratories mainly returned qualitative results for the NAT-based assays, with very few data returned for quantitative assays. As a consequence, some variability was expected in the range of log₁₀ PCR-detectable units/ml, and indeed was found to be comparable to that seen in previous studies carried out to establish other International Standard for NAT-based assays for the blood borne viruses such as HCV [13‐17]. With the exclusion of outlier results, the ranges seen in the other studies are typically between 1.5 and 2.5 log₁₀ PCR-detectable units/ml. Whilst many of assays targeted the 18S rRNA gene, other regions of the P. falciparum genome were used for analysis including the small subunit rRNA gene, the β-tubulin gene and the conserved ATS domain of var genes (Table 2). Despite the differences in the genes targeted by the different assays, there is still a wide range of results, as described above, even when the same region of the P. falciparum genome is targeted, such as the multicopy 18S rRNA gene. In the case of Laboratory 3A, results were reported for a PCR assay directed against the 18S rRNA gene, which gave comparable results to laboratory 7 targeting the single copy β-tubulin gene. In a recent study [28], comparing PCR-based detection methods for P. falciparum on a panel of field samples from Nigeria, a range of assays based upon single or multicopy genes gave a range, the sensitivities of which broadly correlated with copy number, however, this is based upon the work of a single group and not a multicentre study. It would appear that even when the same genes are targeted the difference in response observed by the laboratories will be dependent upon the way the assays are designed (e.g. primer and probe sequences) and implemented (reaction mixes, buffer components, performance characteristics and calibration status of thermal cyclers etc.). The availability of an International Standard will help to reconcile differences such as those mentioned above.

To serve as an International Standard, ideally the preparation should be stable for many years and this is the reason that lyophilized preparations are preferred, where residual moisture and oxygen are minimized [27]. From the Arrhenius model for accelerated degradation [24, 25] the estimates of the potency for the samples stored at elevated temperatures can be used to predict the long term stability at various temperatures. As a consequence accelerated thermal degradation studies were carried out. In the initial stability studies performed on the lyophilized candidate AA, it was observed that there was very little degradation at lower temperatures (i.e. +4°C and +20°C). There was some observed degradation to the samples stored at +37°C and +45°C. This may be due to natural degradation expected at these temperatures, but may also be the result of increasing difficulty in resuspending samples stored at elevated temperatures over time. Whilst the results for the stability of candidate AA are not such a good fit with the model, they nevertheless suggest that this preparation is of adequate stability and support the long term use of the material. Moreover, the results of the stability studies show that a short time at elevated temperatures during shipment of the standard at ambient temperatures should not cause unacceptable loss of potency. Studies are continuing in order to monitor the real-time stability of this material.

The discussions outlined above suggest that candidate AA would be the most suitable to serve as the International Standard. It also has the advantage of being based on blood from a naturally occurring P. falciparum infection and would therefore most closely simulate the field situation. However it should still be noted that the results for candidate BB, based upon the P. falciparum 3D7 clone cultured in leucodepleted blood, demonstrated that the potencies of the two preparations were near identical despite their different formulation and that results are commutable for these two preparations. Candidate preparation AA was therefore selected as the International Standard as it demonstrated that the DNA concentration was in a suitable range that would be acceptable for use, it was suitable for use in a wide variety of assays utilized by the study participants and it demonstrated good stability. The preparation was assigned an arbitrary unitage in International Units (IU), unrelated to the level of parasitaemia or DNA copy number. These arbitrary units can be used to compare assays or laboratory performance by the response given for the International Standard.

Since the introduction of the 1^st WHO International Standard HCV RNA International Standard ten years ago, the number of quantitative assays available has greatly increased. In a recent study to replace the 2^nd WHO International Standard for HCV RNA for NAT-based assays [29] there was a marked increase in the number of quantitative assays and very good agreement with the determinations of the potency of the current standard and the candidate replacements. Such standards can be used for quality control and in the determination of the analytical sensitivity of different testing procedures. They also provide a source of material for assay validation and the production of secondary standards to be used as run controls or working reagents. Indeed the data generated in this present study, produced from a large number of laboratories, using a variety of different assays for P. falciparum DNA detection, demonstrate the wide range of responses to the panel of candidate standards and reinforce the need for a common standard to allow inter-laboratory comparisons of both in-house and commercial assay sensitivities by traceability to a "gold standard" with a physical existence.

In October 2006 the ECBS established this preparation as the 1^st WHO International Standard for Plasmodium falciparum DNA NAT-based assays. This standard has been assigned a unitage of 10⁹ IU/ml. (5 × 10⁸ IU per vial, to be reconstituted in 0.5 ml). The 1^st WHO International Standard for P.falciparum DNA NAT-based assays is stored at the National Institute for Biological Standards and Control (NIBSC code number 04/176). Details on requesting the standard are available on the NIBSC website.