Abstract
QFT Plus variability of 9.6% vs 18.25% in QFT Gold: this could reduce TB reversion/conversion in low-risk individuals http://ow.ly/W97W300mUtZ
To the Editor:
We welcome the findings by Barcellini et al. [1] in their recent publication on the first independent evaluation of QuantiFERON-TB Plus (QFT-Plus; QIAGEN GmbH, Hilden, Germany). The authors conclude that the addition of novel peptides, aimed at stimulating a CD8+ T-cell response, resulted in a diagnostic sensitivity for culture-confirmed tuberculosis of 88% suggesting that the QFT-Plus might offer improved sensitivity when compared with the most recent meta-analysis for QuantiFERON-TB Gold In-Tube (QFT-GIT) [2]. The findings also reinforce recent evidence supporting the theory that CD8+ T-cells have a major role in host defence by stimulating the production of interferon-γ and other soluble factors, activating macrophages which in turn suppress the growth of Mycobacterium tuberculosis (MTB) [3]. We agree with the statement of Barcellini et al. [1] on the need to corroborate such findings and assess sensitivity in CD4+ T-cell depleted patient populations, as in HIV-infected patients where it has been shown that CD8+ T-cells associate with active tuberculosis (TB) [4].
While recognising the importance of providing further insight into the best use of an interferon-γ release assay (IGRA) related CD8+ signal in defining the test's role in programmatic TB infection management [5], we feel that test reproducibility ought also to be assessed. Within subject IGRA variability has been reported in recent research [6, 7]. Such variability could affect test reproducibility and possibly explain the occurrence of apparent conversion/reversion phenomena, particularly in serial testing of low-risk individuals such as healthcare workers.
To assess any potential change in precision of the novel test, as part of an in-house validation exercise, we compared within subject variability between the QFT-Plus and QFT-GIT. As part of the validation of the QFT-GIT, we collected five samples from two donors (10 samples in total) of known MTB status (subject 1: positive and subject 2: negative) via direct collection of blood into the relevant tubes (nil, antigen and mitogen). We sampled the same two subjects for in-house validation of the new QFT-Plus and collected 10 samples from each subject (20 samples in total) into the relevant tubes (nil, TB1, TB2 and mitogen). Samples were well mixed and incubated at 37°C for 16–24 h, then centrifuged at 2500×g for 15 min. Two separate Grifols Triturus automated ELISA analysers (Grifols UK Ltd, Cambridge, UK) (referred to as Triturus A and B) were used to perform the assay sequence at room temperature. The data produced were analysed using QFT-GIT analysis software V2.62 and QFT-Plus analysis software V2.70.
All results for both assays were either positive or negative and no indeterminate results were demonstrated. In both the QFT-GIT and the QFT-Plus, the data for subject 1 expressed 100% positive results indicating likely MTB infection, whereas subject 2 demonstrated 100% negative results indicating no MTB infection. The coefficient of variation (CV) of each batch of subjects on each analyser was calculated based on the mean value for TB antigen/TB1/TB2 minus the nil. With the QFT-GIT, subject 1 expressed a CV for TB antigen−nil of 17.27 and 19.22 on Triturus A and B, respectively, whereas with the QFT-Plus the CVs of TB1/TB2−nil were 10.75/8.10 and 10.98/8.55 on Triturus A and B, respectively. Subject 2, however, produced no reaction therefore there was little variation in the CV of TB antigen/TB1/TB2 minus the nil, with both the QFT-GIT ELISA and the QFT-Plus ELISA expressing a CV of 0.00–0.01 (figure 1).
The objective of this study was to compare the precision of the new QFT-Plus with the QFT-GIT. From the data obtained, qualitatively there is no variation between the two assays; however, when looking at the in-batch data, a lower CV for the QFT-Plus is detected where there has been a positive reaction, as observed in subject 1.
The data obtained from the new QFT-Plus ELISA gave a mean CV of 9.60% in comparison with a mean CV of 18.25% for the QFT-GIT ELISA. The reduction in CV could be particularly important for results around the cut-off point of 0.35 IU·mL−1, by reducing the chance of finding of reversion/conversion phenomena in low-risk individuals where low prevalence decreases the positive predictive value of the test. This research has been conducted within a small target population as an in-house confirmation of the research produced by the manufacturer [8] and presented in the product package insert. While this is useful for in-house validations, larger studies are needed with enough statistical power to allow full assessment of the new assay's precision in the context of QFT-Plus result variability. This, along with assessment of sensitivity and predictive value for incident TB, will provide the necessary body of evidence to guide clinical and programmatic use of QFT-Plus within the context of TB infection management and TB elimination policy [9, 10].
Footnotes
Support statement: QIAGEN GmbH, Hilden, Germany provided the QuantiFERON-TB Plus kits free of charge.
Conflict of interest: Disclosures can be found alongside this article at erj.ersjournals.com
- Received March 23, 2016.
- Accepted April 29, 2016.
- Copyright ©ERS 2016