Evaluating preanalytical sample storage parameters for nucleic acid-based detection of Neisseria gonorrhoeae

Objectives To ensure accurate diagnosis of infectious diseases, preanalytical factors should be considered when assessing specimen quality and subsequent test result. Accordingly, we aimed to systematically assess the effect of storage time, temperature and buffer on the analytical sensitivity of detecting the sexually transmitted pathogen, Neisseria gonorrhoeae across multiple molecular diagnostic platforms. Methods Cultured N. gonorrhoeae was spiked into generic and commercial storage buffers and stored at four temperatures and five time points, ranging from −20°C to 37°C, over 30 days. Samples were processed using the Alinity m STI, Xpert CT/NG and Aptima Combo 2 nucleic acid amplification assays and an in-house quantitative PCR assay. A reduction in analytical sensitivity was defined as a significant (p<0.05) increase in cycle threshold (Ct) value relative to control samples. Results In total, 2756 samples were processed, with N. gonorrhoeae detected in 99.2% of samples. With respect to time, analytical sensitivity was maintained from day 2 (113/120; 94.2%) up to day 30 (110/120; 91.7%) relative to baseline samples. With respect to temperature, analytical sensitivity was maintained from −20°C (147/150; 98.0%) up to 37°C (136/150; 90.7%) relative to baseline samples. Generic buffers, Viral Transport Medium and Amies Liquid Media showed a reduction in analytical sensitivity compared with their commercial counterparts, Aptima Multitest Swab Transport Media and Abbott Alinity transport buffer using select diagnostic assays; this reduction appeared temperature dependent, with the largest differences in median Ct values observed at 37°C (p<0.05). Conclusions Increased prevalence of sample self-collection for sexually transmitted infections (STIs) warrants an evaluation of preanalytical sample storage variables on diagnostic testing performance. Here, across a range of time points, temperatures and storage buffers, N. gonorrhoeae was successfully detected, supporting flexibility in sample storage, and by extension the feasibility of analysing self-collected samples to improve access to STI testing.


INTRODUCTION
Globally, over 1 million sexually transmitted infections (STIs) are acquired daily, including gonorrhoea, which in 2020 caused 82.4 million new cases. 1Considered a curable STI, gonorrhoea is transmitted by the gram-negative, diplococcus, human obligate bacterium Neisseria gonorrhoeae.
With over 50% of all STIs potentially asymptomatic, regular STI check-ups are necessary for early intervention and treatment of STIs. 1 In recent years, sample self-collection for STI testing outside the clinic has become increasingly prevalent, accelerated during the COVID-19 pandemic, and in alignment with global efforts to increase STI diagnostic testing. 2 Accordingly, to ensure accurate and high-quality diagnosis, diagnostic laboratories must consider how preanalytical parameters (eg, sample storage and transport conditions) impact testing performance.With diagnostic platforms providing a presence/absence readout, subtle changes in testing sensitivity may be missed.Previous studies have shown that N. gonorrhoeae sample integrity is best maintained when refrigerated or frozen; however, there are limited data describing how other preanalytical parameters may affect sample integrity, and subsequent test performance in detecting N. gonorrhoeae. 3 4In this study, under controlled laboratory settings, we aimed to systematically assess the effect of storage time, temperature and buffer on the

WHAT IS ALREADY KNOWN ON THIS TOPIC
⇒ Despite the increased prevalence of sample self-collection for sexually transmitted infections (STIs), comprehensive evaluation of pre-analytical sample storage variables on diagnostic testing performance has not been conducted.

Short report
analytical sensitivity of N. gonorrhoeae detection across molecular diagnostic platforms.

Study design and sample preparation
The well-characterised, laboratory N. gonorrhoeae FA1090 strain was cultured for 48 hours at 37°C on chocolate agar in triplicate.To prepare test samples with defined starting concentration, suspensions of cultured N. gonorrhoeae were prepared at 3 McFarland standards using the VITEK DensiChek (BioMérieux, Marcy-l'Étoile, France) instrument according to manufacturer's instructions, with phosphate-buffered saline used for resuspension.Suspensions were then pelleted by centrifugation (16 000×g for 2 min), the supernatant removed and pellets resuspended in  1).

Sample processing
Samples were processed for N. gonorrhoeae detection using the Alinity m STI assay (Abbott, Chicago, Illinois USA) on the Alinity m System, Aptima Combo 2 assay (Hologic, Marlborough, Massachusetts, USA) on the Panther System and Xpert CT/NG assay (Cepheid, Sunnyvale, California, USA) on the GeneXpert System (figure 1); each in accordance with manufacturer's recommendations.Additionally, genomic DNA (gDNA) extracted using the EZ1&2 Virus Mini Kit v2.0 (QIAGEN, Hilden, Germany) on the EZ1 Advanced XL platform was used for an in-house quantitative PCR (qPCR) assay performed on the QuantStudio 5 (ThermoFisher Scientific, Waltham, Massachusetts, USA) targeting the N. gonorrhoeae specific, single-copy porA pseudogene (figure 1; online supplemental table S1). 5

Data analysis
Prism V.9.4.0 was used for all statistical analyses performed.Baseline samples were processed on day 0 at 25°C.Two-way repeated measures analysis of variance (ANOVA) (Šídák correction model) was used to determine the effect of time and temperature, with adjusted p values used to account for the interaction between experimental parameters.In this study using a laboratory strain under controlled conditions, a reduction in microbial load detection is defined by a significant increase in cycle threshold (Ct) value relative to baseline samples (p value <0.05) and referred to as a reduction in analytical sensitivity.This differs from diagnostic sensitivity which typically refers to the ability to detect a pathogen, presence/absence from clinical samples. 6Unpaired t-tests were performed to compare median readout values for analytical sensitivity between storage buffers (p value <0.05) and substantiated using 95% CIs of the median.Invalid and/or

Short report
failed tests to instrumental or sample processing errors were excluded from data analyses.
All buffers were evaluated using an in-house qPCR assay for the porA target gene.Consistent with the other assays, a temperature-dependent difference in buffer performance was observed at 37°C for high and middle concentration samples (online supplemental figures S7D and S8D); high concentration samples stored in VTM (median Ct value 28.9; 95% CI 28.1 to 29.8, n=15) had a significantly higher (p<0.05)median Ct value compared with Aptima Multitest Swab Transport Media (median Ct value 26.2; 95% CI 25.8 to 26.8, n=15), Abbott Alinity transport buffer (median Ct value 26.3; 95% CI 26.0 to 27.4, n=15) and Amies Liquid Media (median Ct value 25.7; 95% CI 25.1 to 26.3, n=15) (online supplemental figure S7D).However, contrary to the other assays performed, significant differences between buffers were not observed at 37°C for low concentration samples processed using qPCR for the porA target gene (online supplemental figure S9D); most likely attributable to several VTM samples (8/15; 53.3%) exceeding qPCR limit of detection under these testing parameters.Overall, these results demonstrate that while subtle reductions in analytical sensitivity were observed for select buffers, there is certainly scope for context-specific flexibility in storage buffer selection.

DISCUSSION
To improve access to diagnostic services, the self-collection of patient samples has become increasingly prevalent, offering a wide-reaching and novel framework for gonorrhoea diagnosis.However, limited emphasis has been placed on identifying the acceptable limits of parameters required for maintaining N. gonorrhoeae sample integrity during sample collection, storage and transport to ensure accurate downstream test performance.Here, our results extend on current literature and highlight

Short report
that service likely deviate from stringent sample storage standards to ensure continuity in service delivery; particularly important during consumable shortages and for outbreak preparedness. 3 4In saying that, the reduced performance of generic storage buffers observed in this study may compromise downstream bacterial culture and/or next-generation sequencing, for which sufficient microbial load is required for a fully informative analysis 7 ; although this was not tested in our study.In addition, our findings suggest that VTM should not be used for samples processed using the Xpert CT/NG assay.Importantly, our findings were applicable across a range of routinely used diagnostic platforms; however, due to instrumental errors caused by the Panther System, trends attributable to the effects of time and temperature were unable to be discerned from the Aptima Combo 2 readouts.In addition, molecular testing for gonorrhoea is often coupled with chlamydia; however, due to the complexity of culturing obligate intracellular bacterium Chlamydia trachomatis in the laboratory, we were unable to examine how sample storage parameters impact the testing sensitivity of C. trachomatis.A limitation of this study is that tested samples lacked a background clinical matrix, and thus cannot be directly extrapolated to the storage of clinical samples.In clinical samples, host and commensal genomic material, including nucleases, and DNA extraction and amplification inhibitors can potentially affect sample integrity and nucleic acid amplification test performance. 8Furthermore, the anatomical site of sample collection may impact on the composition of the clinical matrix and microbial load. 9In conclusion, our data demonstrate that regardless of variations in sample storage parameters, in particular time and temperature, N. gonorrhoeae samples can be accurately detected across molecular diagnostic platforms.This work provides a premise for service providers, and by extension diagnostic laboratories to offer more accessible STI testing, specifically within the self-collection framework.

Figure 1
Figure 1 Schematic flow chart illustrating the storage parameters and molecular assays used to assess the effect of storage time, temperature and buffer on the analytical sensitivity of detecting Neisseria gonorrhoeae.ˆAs per manufacturer's recommendations, only samples stored in Amies Liquid Media and Viral Transport Medium were processed with the Xpert CT/NG assay.# As per manufacturer's recommendations, only samples stored in Aptima Multitest Swab Transport Media were processed using the Aptima Combo 2 assay.Created with BioRender.com.Aptima, Aptima Multi Swab Transport Media; Alinity, Abbott Alinity transport buffer; Amies, Amies Liquid Media; VTM, Viral Transport Medium; gDNA, Genomic DNA.

of storage buffer on analytical sensitivity of N. gonorrhoeae detection
).With respect to temperature, analytical sensitivity was highest for samples stored at −20°C (147/150; 98.0%) and 4°C (147/150, 98.0%).However, a relatively high analytical sensitivity was also observed for samples stored at 25°C (139/150; 92.7%) and 37°C (136/150; 90.7%) (figure2; online supplemental figure S1-S9); reductions in analytical sensitivity observed at these temperatures are likely interrelated to the other variables tested.Overall, these results demonstrate that in a controlled laboratory setting, flexibility in sample storage can be afforded for up to 30 days and 37°C.Effect