Abstract

Background: Most of the samples sent to reference laboratories are delivered by post. Thus, diagnostic PCR tests on blood samples have to be performed using methods which are optimised and validated for such conditions. There is a low probability that the organisms Toxoplasma gondii and Borrelia burgdorferi will be present.

Aim: To confirm that robotic extraction methods followed by real time PCR will detect as little as one organism/test sample in postal specimens.

Methods: Human blood samples spiked with decreasing numbers of each organism (range 10⁵–1/per extract) were extracted using two commercial kits on a Qiagen BioRobot EZ1 Workstation. Extracts of whole blood and blood fractions were tested by real time PCR. The effect of storage of blood for 1–6 days at room temperature was also investigated.

Results: Maximum sensitivity (1 organism/test sample) was achieved for T gondii with either extraction method; the sensitivity for B burgdorferi was between 1 and 10 organisms/test. Whole blood was the most suitable sample to extract, as both organisms were as likely to be detectable in the red cell as the white cell fraction. Sensitivity was not reduced by storing spiked samples at room temperature for up to 6 days. Inhibitory effects on PCR were not a significant problem provided that samples were extracted using the blood extraction kit.

Conclusions: Using appropriate robotic extraction methods, both T gondii and B burgdorferi can be detected by real time PCR with near maximum possible sensitivity in whole blood samples. Blood samples can be transferred to reference laboratories by post without loss of sensitivity over the likely transit period.