Use of saliva-based qPCR diagnostics for the accurate, rapid, and inexpensive detection of strep throat

Introduction

Antibiotic resistance is a global public health concern, and the World Health Organization urges countries to take action against this threat at the national, regional, and local level [1]. The United States National Action Plan for Combating Antibiotic-Resistant Bacteria calls for the development and validation of new diagnostics to rapidly distinguish between viral and bacterial pathogens and for these diagnostics to be easily implemented [2]. In addition, the Centers for Disease Control and Prevention calls for improvement of antibiotic prescribing to minimize missed and delayed diagnoses, placing viral pharyngitis and strep throat (ST) as top priority conditions [3]. Outpatient health care facilities were responsible for prescribing approximately 269 million antibiotic prescriptions in 2015, and it was estimated that 30 % of those prescriptions were inappropriate [4, 5]. Pharyngitis is commonly diagnosed in outpatient settings and is the third top diagnosis for which antibiotics are prescribed [5]. In adults with pharyngitis, 10 % have ST, the only form of pharyngitis requiring antibiotics [6]; however, approximately 60 % of adults with pharyngitis are prescribed antibiotics [7]. These statistics highlight the pervasiveness of the inappropriate prescribing of antibiotics for pharyngitis and suggest that antibiotics are being prescribed even with a negative ST diagnostic result. Outpatient health care facilities currently utilize a rapid antigen detection test (RADT) for ST, but those screening procedures often yield a high rate of false negative results. The most common RADT is the lateral flow immunoassay, whereby antibodies detect group A Streptococcus (GAS) cell wall antigens from throat swab samples [8]. A 2014 meta-analysis analyzing 48 studies found that the sensitivity (true-positive rate) of ST RADTs averaged 86 % (with lateral flow immunoassays ranging from 59-96 % sensitivity) and the specificity (true-negative rate) averaged 96 % [8]. Therefore, many false negative diagnoses can occur, depending on the brand of RADT used and the proficiency of those performing the test [9]. It is best practice to verify a negative RADT diagnosis with a reference standard beta-hemolytic GAS culture; however, this takes 24–72 h to grow and is an additional cost, so many clinics do not perform this secondary test. Unfortunately, a false negative result can be detrimental to patient health in that it can delay recovery and can lead to further complications when antibiotics are not prescribed [10]. In one study, 42 % of patients with a false negative RADT diagnosis had moderate to heavy bacterial burden, and more than 4 % suffered from severe complications associated with untreated ST [10]. Although widespread use of the ST RADT has reduced the overall number of antibiotic prescriptions for pharyngitis [11], and average ST RADT specificity is relatively high, specificity for these tests can still range from 78–100 %. This can result in false positive diagnoses, which further contributes to the unnecessary prescription of antibiotics [12]. Altogether, there is great need for more sensitive and specific ST diagnostic tools in outpatient settings.

One of the most sensitive diagnostic tools used to detect viral or bacterial DNA in a sample is quantitative real-time polymerase chain reaction (qPCR). This technology has been tested rigorously in diagnosing hundreds of diseases and is generally more sensitive and specific compared to other diagnostic tests [13]. Uhl et al. analyzed qPCR detection of GAS vs. the RADT (compared to beta-hemolytic GAS culture, the reference standard), and demonstrated that qPCR was more sensitive than the RADT (93 vs. 55 %, respectively) and both had comparable specificities (qPCR 98 % vs. RADT 99 %) [14]. Although qPCR can be an accurate diagnostic tool, this technology has historically been costly, time-consuming, and has required specialized training to perform; therefore, it has not commonly been used as a diagnostic tool in outpatient clinics. Within the last decade, next generation qPCR technology has emerged (smaller, faster, easier to operate and more cost effective), providing the potential to overcome traditional barriers to using qPCR for the diagnosis of common infections in outpatient health care facilities.

The aim of this research was to test the efficacy and speed of a novel saliva-based ST diagnostic protocol to establish a new low-cost, accurate diagnostic tool that has the potential to be used in outpatient health care facilities. Two key innovative elements highlighted in this research are: (1) development of a diagnostic protocol for ST relying solely on saliva samples (compared to traditional testing with a throat swab) and (2) use of the low-cost, rapid Open qPCR system (Chai Biotechnologies, Santa Clara, CA, USA) as a potential diagnostic tool. Saliva sample collection is inexpensive and non-invasive; however, saliva is an under-utilized tool in diagnosing disease and its collection is not standardized. Research shows that that GAS can be readily detected via qPCR in the saliva of infected patients [13, 15] but these protocols are lengthy, expensive, and require specialized training to carry out the diagnostic tests. To test saliva samples, we created a simple Chelex 100-based DNA extraction technique, and optimized a fast qPCR protocol (40 cycles in 20 min) for the accurate, rapid and inexpensive detection of ST.

Materials and methods

Results

DNA extraction from saliva samples

Various DNA extraction protocols were compared and tested in the process of developing a low-cost, simple DNA extraction step for this study. Compared to complex, expensive protocols used by standard DNA purification kits [17], and long protocols involving the more cost-effective Chelex 100 resin [18, 19], we developed a Chelex-based DNA extraction protocol for saliva samples that involves six simple steps, takes 2 min to perform, and costs approximately $0.07 per sample to process (Figure 1).

Figure 1:

Comparison of DNA extraction protocols. DNeasy Blood & Tissue Kit (Qiagen, Germantown, MD) protocol and cost information were gathered from the DNeasy Blood & Tissue Saliva Protocol [17] and online pricing, respectively. Full and Simplified Chelex 100 protocols were found in the literature [18, 19], and average total costs were calculated based on the average online pricing of equipment and our pricing of Chelex 100.

When optimizing our DNA extraction protocol for saliva samples, we tested many variables: vortexing vs. shaking by hand, boiling at 100 °C for 10 min vs. a higher temperature for a shorter period of time, and length of centrifuge spin (Table 1). Figure 2 demonstrates the importance of adding 10 % Chelex 100 and a boiling step (1 min at 120 °C) for sufficient DNA extraction and GAS DNA amplification via qPCR. Not only did addition of 10 % Chelex 100 plus a boiling step allow for a more robust detection of GAS DNA (i.e. lower quantification cycle (Cq)) (Figure 2A), but it was a critical step for the detection of GAS DNA in some samples (Figure 2B).

Table 1:

Protocol testing and optimization steps.

DNA extraction testing
Vortexing vs. shaking by hand	Average Cq
No vortexing or shaking by hand	35.25
Shaking by hand once	37.29
Shaking by hand twice	32.61
Vortexing twice	33.29
Sample boiling temperature and time	Average Cq
100 °C for 10 min	33.38
100 °C for 5 min	34.03
110 °C for 5 min	32.13
120 °C for 3 min	28.30
120 °C for 2 min	31.96
120 °C for 1 min	30.50
120 °C for 45 s	36.03
120 °C for 30 s	34.61
Length of centrifuge spin	Pellet formed
5 s	Yes
10 s	Yes
15 s	Yes
20 s	Yes
qPCR protocol testing
Protocol speeda	Average Cq
Standard protocol	33.29
Fast protocol 1	34.41
Fast protocol 2	38.10
Fast protocol 3	36.07

1. ^aA total of 20 fast protocols were tested. Data from three fast protocols are displayed here. Bolded, text indicates steps that were selected for the final protocol. Cq, quantification cycle.

Figure 2:

qPCR detection of group A Streptococcus DNA was enhanced in saliva samples processed with 10 % Chelex 100 plus a 1 min boil at 120 °C when run on a fast qPCR protocol. (A) Strep positive sample A was treated with either no 10 % Chelex 100 and no boiling step (quantification cycle (Cq)=34.49), 10 % Chelex 100 and no boiling step (Cq=33.69), or 10 % Chelex 100 and a 1 min boil at 120 °C (Cq=32.77). (B) Strep positive sample B was treated with either no 10 % Chelex 100 and no boiling step (Cq=no amplification), 10 % Chelex 100 and no boiling step (Cq=no amplification) or 10 % Chelex 100 and a 1 min boil at 120 °C (Cq=34.16).

Open qPCR ST diagnostic protocol development

To develop an Open qPCR diagnostic protocol for ST, we tested various qPCR protocols and GAS primers. We began the protocol optimization process by running a standard 53+ minute qPCR protocol (95 °C for 10 min followed by 40 cycles of 95 °C for 15 s, 55 °C for 22 s, 72 °C for 30 s) on the CFX Connect Real-Time PCR machine (Bio-Rad Laboratories, Berkeley, CA, USA). Expression of the 16s bacterial housekeeping gene was analyzed in preliminary testing of saliva samples (Supplemental Figure 1). Three different GAS primers specific for the dnaseB gene (1 primer set) and speB gene (2 primer sets) were tested using DNA extracted from participant saliva. One speB primer set displayed the highest sensitivity and specificity for GAS detection in all samples and was used hereafter in this study. In order to identify the optimal parameters to amplify GAS DNA with speed and accuracy, samples underwent repeated testing on both the CFX Connect Real-Time PCR and Open qPCR machines under 20 different fast protocols with slightly altered denaturation, annealing and extension times (Table 1). Our testing identified the following qPCR protocol as being optimal: initial polymerase activation/DNA denaturation step at 95 °C for 30 s followed by 40 cycles of denaturation at 95 °C for 4 s and annealing/extension at 67 °C for 18 s. Our final, optimized saliva-based Open qPCR ST diagnostic test protocol (Figure 3) can be run significantly faster than a standard qPCR protocol, with positive results developing at 20 min and negative results confirmed by 24 min. Diagnostic results are made clear with ST positive samples showing the presence of an upward sloping line, indicating GAS DNA is present in the sample, and ST negative samples showing the presence of a flat line, indicating GAS DNA is not present in the sample (Figure 4).

Figure 3:

Final optimized saliva-based Open qPCR strep throat diagnostic test protocol. ST, strep throat; GAS, group A Streptococcus.

Figure 4:

Example Open qPCR diagnostic readout. Samples in wells A1–B1 were strep throat positive (ST +) samples that displayed group A Streptococcus (GAS) DNA amplification, indicated by an upward sloping line and a calculated quantification cycle (Cq) value. Samples in wells B2–B8 were ST negative (ST −) samples that displayed no GAS DNA amplification, indicated by a flat line and no calculated Cq value.

Discussion

The results of this study demonstrate that a simple Chelex 100 protocol can be used to quickly extract GAS DNA from saliva and that longer Chelex 100 protocols are not required for effective downstream amplification of GAS DNA. Chelex 100 resin was used in this study due to its low cost and ability to bind to multivalent metal ions to inhibit DNases from degrading DNA after the critical boiling step (Figure 2) [20]. Our data additionally show that primers specific for the speB gene, when run through a fast qPCR protocol (40 cycles in 20 min), are capable of accurately diagnosing ST with 100 % sensitivity and 100 % specificity compared to reference standard diagnostics (Table 2). Not only is saliva collection less invasive than the standard throat swab, DNA in saliva samples can remain stable for months [21] and can be extremely effective in diagnosing bacterial and viral illnesses via qPCR, as demonstrated in this study and others [13, 15].

Although RADTs can provide results within 2–9 min (Table 3), they exhibit low sensitivity which leads to inappropriate treatment for those who receive false negative results. The potential also exists that heath care providers, knowledgeable of this low sensitivity, may prescribe antibiotics even when a patient has a negative test result; this would account for the recorded overprescription of antibiotics in patients with pharyngitis [7]. Although the Open qPCR ST diagnostic protocol takes slightly longer than RADTs to produce results (22–26 min, Table 3), it can provide 100 % sensitivity, specificity and accuracy in diagnosing ST (Table 2). Additionally, up to 16 tests can be run concurrently in the Open qPCR machine, allowing for batch testing to be performed. No special qPCR training is required to run the Open qPCR ST diagnostic test protocol (Figure 3), and test results are easy to read (Figure 4). Following the one-time purchase of an Open qPCR machine, a heat block, and a mini centrifuge (approximately $5,000 in total), samples can be processed and run through the full protocol for approximately $1.12 per sample. This is a fraction of the cost and time it would take to run qPCR diagnostics at an external lab or perform beta-hemolytic GAS culture. Additionally, we found that 10 % Chelex 100 and a ST qPCR Master Mix (see Materials and methods) could be prepared in advance and stored at 4 °C and −20 °C, respectively, for up to at least 5 months (i.e. the duration of our studies). Preparing these reagents in advanced allowed for saliva samples to be tested immediately as soon as they became available.

We acknowledge that there are barriers to implementation of this diagnostic protocol in the clinical setting. A larger dissemination and implementation study would identify these barriers and serve to spread knowledge and encourage integration of this new diagnostic protocol. Widespread adoption of this diagnostic test would require the following: (1) obtaining a Clinical Laboratory Improvement Amendments (CLIA) waver for this test to be used in CLIA-certified labs, (2) making prepared test reagents commercially available (i.e. Tubes A & B, Figure 3) and (3) utilization of a qPCR machine approved for diagnostic procedures. From the perspective of health care practitioners and administrators, additional barriers to implementation may exist. These barriers may include resistance to change and reluctance to adopt new protocols and procedures, plus possible budgetary constraints surrounding the cost of purchasing equipment (approximately $5,000). However, the benefits of this diagnostic test are clear. With 100 % sensitive and 100 % specific results, follow-up GAS cultures will no longer be required to confirm a negative diagnosis. Antibiotics will be prescribed appropriately, due to the lack of false negative or false positive results. Saliva sample procurement is simple, noninvasive, and preferable among patients. DNA extraction and qPCR testing are quick and easy to perform, producing results in 26 min or less at a minimal cost.

Our results demonstrate that saliva-based Open qPCR diagnostics yield rapid, accurate, cost-effective test results for ST, thus making it a suitable platform for outpatient health care facilities. This study could lay the groundwork for future Open qPCR testing for the identification of other bacterial, viral, or parasitic pathogens known to be detectable in saliva using PCR or qPCR [15]. Our 2 min saliva DNA extraction method, in combination with a fast qPCR protocol, could be optimized to detect bacterial DNA such as Streptococcus pneumoniae (causative agent of pneumonia) [22], Mycobacterium tuberculosis (causative agent of tuberculosis) [23], Treponema pallidum (causative agent of syphilis) [24], and Helicobacter pylori [25]. Diagnostics could similarly be developed to identify viral DNA such as Epstein-Barr virus (causative agent of infectious mononucleosis) [26], hepatitis B [27], herpes simplex virus [28], cytomegalovirus [29], and human herpesvirus [30], as well as parasitic DNA such as Plasmodium falciparum (causative agent of malaria) [31]. Although PCR and qPCR protocols do exist to detect all of these pathogens in saliva, these protocols are long, expensive, and require specialized training to perform. The rapid diagnostic protocol presented in this study could provide the framework for the development of fast, accurate tests for these common infectious diseases.