Development of a Triplex Endpoint PCR Assay for the Detection of SARS-CoV-2: Insights on Cost-Efficiency and Method Design

Lower respiratory tract infections make up a significant proportion of all respiratory infections, being related to substantial mortality, morbidity and economic burden. Viruses are described as responsible for 10–60% of the lower respiratory tract infections, depending mostly on the socio-demographic index, thus ranking on the first places next to Streptococcus pneumoniae, Staphylococcus aureus or Mycoplasma pneumoniae [1‒3]. It is important to determine the cause of infections, especially in the actual global context, and particularly in the winter season, when the etiology can include not only SARS-CoV-2 but also Influenza A/B or syncytial respiratory virus [4]. Even though some symptoms are comparable for both types of infections, there are many differences, and the clinical presentation for COVID-19 can be severe [5].

Among viral respiratory tract infections, SARS-CoV-2 infection is currently one of the most surveyed pathological conditions. COVID-19 has severely affected the world population from various perspectives: healthcare systems were overburdened, the industry suffered a backlash, and behavioral health problems and educational consequences have been constantly reported [6‒8].

The development of diagnostic tests for the detection of SARS-CoV-2 has become essential at the global level. On this occasion, we realized the importance of developing new diagnostic tests to rapidly identify infected persons, isolate patients, and prevent the spread of the disease [9]. This does not apply only to SARS-CoV-2, but to any other potential situation that may arise in the future. Moreover, a diversity of tests is needed, as the laboratories around the world are different in terms of infrastructure and performance.

According to the statistics, the highest number of tests performed in human history was conducted to detect SARS-CoV-2 infection [10]. Nowadays, several types of tests are used for this purpose, including molecular, antigen, and antibody tests. Each test type has its strengths and limitations, and choosing the appropriate one depends on the purpose, timing, prevalence of the virus in the population, and laboratory facilities [11]. The need for good diagnostic tools was essential from the beginning, and more than 6 billion tests for COVID-19 were performed during the pandemic, with qRT-PCR being the gold standard of diagnosis [12]. Diagnostic strategies have shifted from a pandemic response to infection control [11], and diagnostic tests will continue to be important, as COVID-19 continues to affect the human population.

Besides having multiple advantages, qRT-PCR also has a series of limitations [13], an important one being the relatively high price (USD 25–100); also, the requirement of relatively expensive equipment might limit access to this diagnostic method, especially in healthcare settings with limited resources [14]. Also, due to the constant virus mutations that appear especially in the S gene, some commercial qRT-PCR diagnostic kits showed gene target failure, which is a double-edged sword: can be used as a differentiation tool for virus variants [15] but can also mislead the diagnostic result. To address this issue, periodic adjustments of the diagnostic kits are required, particularly by refining the structure of the primer sets, which are responsible for recognizing the specific viral gene sites, in adaptation to emerging virus variants.

The purpose of the study was to build a new, highly accurate, fast, and efficient SARS-CoV-2 PCR-based detection method, with lower costs than the current methods, applicable in limited settings, and which can successfully detect all current variants. Implementation of such tests could facilitate a large-scale diagnosis of SARS-CoV-2 infections, enabling rapid and appropriate therapeutic management for each case and ongoing monitoring of patients. In this context, expanding the availability of cost-effective and accurate diagnostic tools is critical for diagnosing infectious diseases that pose significant global health threats, especially in low-resource settings where access to advanced medical technologies is often limited. As viruses continue to evolve and new infectious agents emerge, the ability to quickly adapt diagnostic methods is essential. Moreover, being a diagnostic test based on primer design, the thorough description of the method allows for easy adaptation to detect other pathogens, thereby supporting the need for continuous research and innovation in diagnostic technologies to ensure preparedness for future pandemics.

The study was conducted in the molecular biology department of Mures County Clinical Hospital (MCCH) from Târgu Mureș, between 2020 and 2022. The study was approved by the Ethical Board of MCCH (number 8665/May 09, 2023), allowing the frozen RNA samples from COVID-19 patients to be used for further experiments.

The pharyngeal and nasal secretions were collected by respecting the internal protocols and transported on VTM (Virus Transport Medium, Sanimed, Romania) to the molecular laboratory. The samples were collected from outpatients (via the Public Health Department) and from hospitalized patients suspected of SARS-CoV-2 infection.

The routine IVD diagnostic for SARS-CoV-2 was performed in the molecular biology laboratory, using the real-time qRT-PCR method. For this, the viral RNA was extracted using the Quick-RNA Viral Kit (Zymo Research, U.S.A.) or with Genolution Nextractor NX-48S (Genolution, Korea) automatic extractor, following the manufacturer protocol. From the obtained RNA, 5 μL were used for PCR amplification, and the remaining volume was stocked at −70°C for further tests. The IVD qRT-PCR was performed in the QuantStudio 5 Real-Time PCR equipment (Thermo Fisher Scientific, Singapore), using the TaqPath COVID-19 RT-PCR IVD diagnostic kit (Thermo Fischer Scientific, Waltham, MA, USA), following the presence of 3 specific viral genes (N, ORF, S).

The selection criteria for the samples to be further used for our assay development were only the first sample for each patient with a Ct (cycle threshold) value under 35 for the ORF gene and a minimum of 2 genes detected by the reference method using the TaqPath COVID-19 RT-PCR IVD diagnostic kit. From the stocked RNA, 92 samples were randomly selected from the main peaks of the COVID-19 infection periods, to prove the functionality of the assay on the different SARS-CoV-2 variants: 32 samples from November 2020, 30 from April 2021, and 30 from February 2022. Also, out of the pandemic period, 30 recent samples (September–October 2023) were analyzed. Negative controls consisted of 36 random negative samples as proved by the reference IVD PCR method.

For all selected samples, complementary DNA (cDNA) was prepared from total RNA by reverse transcription, using GoScript reverse transcription kit (Promega, USA). The RNA was normalized at 1 μg/µL and treated with DNase I (Thermo Fisher Scientific, Waltham, MA, USA). Each reaction was performed following the manufacturer protocol, in a final volume of 20 μL consisting of 1 μL of RNA, 1 μL of oligo dT, 8.5 μL reaction mix, and 9.5 μL water. The reverse transcription protocol consisted of one cycle at 25°C for 5 min (annealing phase), followed by extension for 1 h at 42°C, with a final step at 70°C for 15 min (inactivation of the reverse transcriptase). The cDNA was quantified using the D30 BioPhotometer (Eppendorf AG, Hamburg, Germany) and stocked at −20°C.

To evaluate the presence and integrity of the extracted RNA, amplification of human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) housekeeping gene was followed by PCR. The amplification mix consisted in 12.5 μL DreamTaq Green PCR Master Mix 2X (Thermo Fisher Scientific, Waltham, MA, USA), 0.2 μm of each forward (5′ GTC​TCC​TCT​GAC​TTC​AAC​AGC​G 3′) and reverse (5′ ACC​ACC​CTG​TTG​CTG​TAG​CCA​A 3′) GAPDH primers and water to a final volume of 25 μL. The amplification protocol consisted of 1 cycle for initial denaturation (95°C for 1 min), 40 amplification cycles (95°C for 30 s for denaturation, 60°C for 30 s annealing, 72°C for 1 min the final extension), and 1 final extension at 72°C for 10 min.

The amplification products were visualized by gel electrophoresis (50 min at 100 V) in 1% agarose gel containing GelRed^® (Sigma‐Adrich, Saint Louis, Missouri, USA), and the molecular size was appreciated against the Thermo Scientific GeneRuler 1 kb DNA Ladder (Thermo Fisher Scientific, Waltham, MA, USA). The final image was captured using MiniBIS Pro (Bio‐Imaging Systems, Modiʹin‐Maccabim‐Reʹut, Israel) equipment, expecting visualization of 131 bp bands after gel electrophoresis after successful GAPDH amplification.

The rationale for selecting the primers was to identify pairs capable of amplifying multiple regions of the SARS-CoV-2 genome in a triplex RT-PCR assay, with high specificity and sensitivity across different virus variants, while avoiding cross-reactions and facilitating straightforward result interpretation, despite potential gene target failures. The design of the primers was performed using the National Library of Medicine (NIH) database, following several recommendations: repetitive sequences should be avoided, the amplicon lengths should differentiate between the target genes, the G (guanine) and C (cytosine) recommended percent is 40–60%, the melting temperature (Tm) of primers should be between 60 and 63°C, self-complementarity should not exceed 8. The primers were designed based on the Wuhan Hu-1 genome (NCBI reference genome NC_045512.2), for 3 specific viral genes (ORF1ab, S, and N). The generated primers were crosschecked with other SARS-CoV-2 genomes on NCBI database: B.1.617.2-Delta (GenBank accession number OK091006.1), B.1.1.529-Omicron (GenBank accession number OW996240.1), and the current variants-of-interest variants such as BA.2.86 (GISAID EPI_ISL_18509742), JN.1 (GISAID EPI_ISL_18735273), XBB (GISAID EPI_ISL_18509732), or EG.5 (GISAID EPI_ISL_18406668). Also, the primers were checked against the human genome (NCBI TaxId 9606), to check potential nonspecific amplifications. The selected primer sequences for the study are presented in Table 1.

The resulting primers were compared across genomes using Primer-BLAST to ensure that the selected sequences could detect all variants of SARS-CoV-2 circulating in Romania during the respective period (B1.1.7-Alpha, BA.1.617.2-Delta, and B.1.1.529-Omicron). The primer sequences were adapted to ensure compatibility with various reaction conditions and their interactions were analyzed in silico for potential cross-reaction (cross-dimers, self-dimers, and hairpins) using Beacon Designer [16] (Table 2).

Simplex PCR. The functionality of the primers was assessed initially by endpoint PCR reaction, using 3 positive samples (as detected through reference method), using the selected primers (Table 1), in a mix consisting of 12.5 μL DreamTaq Green PCR Master Mix 2X (Thermo Fisher Scientific, Waltham, MA, USA), 0.08 μm of each forward and reverse primers and water to a final volume of 25 μL. The PCR was performed under the following conditions: initial denaturation (95°C for 1 min), 40 cycles (denaturation at 95°C for 30 s, annealing at 60°C for 30 s, extension at 72°C for 1 min), and final extension (72°C for 10 min). The amplification products were loaded in 1% agarose gel containing GelRed (Biotium, UK), for 50 min at 100 V. Bands corresponding to 113 bp (for N gene), 228 bp (for ORF1ab gene), and 371 bp (for S gene), respectively, confirmed proper amplification.

Sequencing of Amplification Products. The electrophoresis bands for the amplification products of 2 random samples were excised from the gel, and the amplified DNA was extracted and purified using the Gel/PCR DNA Fragments Extraction Kit (Geneaid Biotech, New Taipei City, Taiwan). The amplicons were further processed using next-generation sequencing with the DNA Prep with Enrichment Kit (Illumina), and the final library was sequenced using the Illumina MiSeq Reagent Kit v3 (600 cycle run, producing 2 × 300 bp paired-end reads). The assembly and mapping of the sequencing reads to the reference sequence were performed using HAVoC, a bioinformatic pipeline developed by the University of Helsinki, installed and used from a command line in Conda environment [17]. Following the assembly, a FASTA consensus sequence resulted for each sequenced sample, which was then aligned with the SARS-CoV-2 reference sequence using Blastn [18].

Serial dilutions of 10⁸, 10⁶, and 10⁴ DNA copies/µL were prepared in nuclease-free water. Using these standards, the primer efficiency was assessed by real-time PCR, in a reaction mix consisting of GoTaq qPCR Master Mix (Promega), 0.1 μm of each forward and reverse primers, and 1 μL of each standard DNA. The amplification was performed in QuantStudio 5 real-time PCR equipment (Thermo Fisher Scientific, Singapore), following the protocol: 1 cycle initial denaturation at 95°C for 2 min, 40 cycles (95°C for 15 s denaturation, followed by annealing and extension at 60°C for 1 min). A final melting curve step was also included. The BRYT green fluorescence was captured after each cycle. The criteria applied for proving the primer efficiency for all three genes were slope −3.3 ± 0.5, amplification efficiency over 80%, with a correlation factor r² between 0.9 and 1.0, and a melting curve with a single peak. For all reactions, a no template control was used (nuclease-free water instead of target DNA).

After the primer validation, the confirmation of the method was finally assessed by triplex PCR. The optimization of the method was performed using different concentrations of primers (Table 1) and cDNA template, establishing the optimal concentrations for a reliable PCR result. The optimal reaction mix consisted in 12.5 μL DreamTaq Green PCR Master Mix 2X (Thermo Scientific, Waltham, MA USA), 0.08 μM of primers (for S and N genes), and 0.16 μM of primers (for ORF1ab gene), 1 μL cDNA, and water to final volume of 25 μL. The amplification protocol was set up as follows: 1 cycle initial denaturation at 95°C for 1 min, 40 cycles (denaturation at 95°C for 30 s, 60°C for 30 s annealing, 72°C for 1 min extension) and a final extension at 72°C for 10 min.

The amplification products were loaded in 2% agarose gel, containing 1 μL of GelRed (Biotium, UK) in TAE buffer for 50 min at 100 mV. The result was visualized in UV spectrum using MiniBIS Pro (Bio‐Imaging Systems, Modiʹin‐Maccabim‐Reʹut, Israel). Negative samples, confirmed by the IVD kit, were also used as negative controls to demonstrate the absence of nonspecific amplifications.

For the implementation of the method in routine diagnostic, a triplex PCR assay was performed using a one-step reaction mix that incorporates both reverse transcription and amplification. For this, a reaction mix consisted of 12.5 μL TaqPath 1-Step RT-qPCR Master Mix (Thermo Scientific, Waltham, MA USA), 0.04 μM of primers (for S and N genes), and 0.08 μM of primers (for ORF1ab gene), 5 μL RNA, and water to final volume of 50 μL. The amplification protocol was set up as follows: 1 cycle of UNG incubation at 25°C for 2 min, 1 cycle of reverse transcription at 50°C for 15 min, 1 cycle of polymerase activation at 95°C for 2 min, 40 amplification cycles (denaturation at 95°C for 15 s, annealing and extension at 60°C for 60 s), and a final extension at 60°C for 5 min. The amplification products were loaded in 2% agarose gel, containing 1 μL of GelRed (Biotium, UK) in TAE buffer for 50 min at 100 mV. The result was visualized in UV spectrum using MiniBIS Pro (Bio‐Imaging Systems, Israel).

The sensitivity and specificity of the developed endpoint RT-PCR were assessed compared to the TaqPath 1-Step RT-qPCR results, using GraphPad InStat 3 software.

A total number of 41,818 samples were analyzed in the molecular laboratory of the MCCH between June 2020 and December 2022. Of these, 26.15% (n = 10,936) were positive for SARS-CoV-2, 70.9% were negative, and 2.95% were inconclusive or invalid. Three time periods with a large number of positive tests (November 2020, April 2021, and February 2022), considered peaks, have been identified in Romania (Fig. 1).

The evolution of positive cases in MCCH fit the Romanian evolution trend, especially in the first wave of COVID-19 (October–December 2019); in the following waves, the COVID-19 prevalence was lower than the national one, but still visible as peaks in the graphical evolution, the peaks from 2021 October and 2022 February being due to the Omicron variant. Compared with the global evolution, the MCCH and national evolution did not fit completely; most positive pandemic cases evolved in January–March 2022, conversely to the evolution in Romania and MCCH, where the most important peak was at the beginning of the pandemic, in October–November 2020. Romania was placed in the last position in Europe regarding the vaccination rate, and the restrictive measures were relaxed in the summer of 2021. These were probably some reasons that led to the important evolution of the third important COVID-19 peak in the autumn of 2021, due to the Delta variant of SARS-CoV-2, conversely to the global evolution of positive cases [20].

From the preserved RNA samples, 32 were selected for this study from 2020 (considered Wuhan wild-type), 30 samples from 2021 (considered Delta variant), and 30 from 2022 (Omicron variant). For all 92 samples, the nucleic acid analysis by spectrophotometry confirmed the presence of RNA. The sampling quality and the integrity of the extracted RNA were confirmed by RT-PCR for the GAPDH human housekeeping gene, as it this is a good extraction and quality marker [21]. All the samples presented amplification products of 131 bp, proving the integrity of RNA (Fig. 2).

The results of the initial primer validation by endpoint PCR showed amplification products at approximately 113 bp for N, 228 bp for ORF1ab, and 371 bp for S genes, confirming the specificity of the primers on the SARS-CoV-2 genome (Fig. 3). The sequencing of amplification products showed a successful match with the target genes for the Wuhan-WH-1 genome with NCBI (accession number NC_045512.2), as shown in Table 3.

The primer efficiency was tested by real-time qRT-PCR (Fig. 4), showing values of 91.22% for N gene (r² = 1.00, slope = −3.552), 84.87% for ORF1ab (r² = 0.998, slope = −3.747), and 81.51% for S gene (r² = 0.999, slope = −3.863). Studies showed that the primer efficiency between 80 and 110% is consistent with a valid method [22].

Figure 5 presents the result for triplex PCR, showing amplifications at specific molecular weights, as expected by the in silico evaluation: 113 bp, 228 bp, and 371 bp for N, ORF1ab, S genes, respectively. One advantage of this method is the easy interpretation of the results, without the need for complex knowledge of qRT-PCR details such as the shape of the amplification curve or cycle threshold values, which can mislead a laboratory practitioner [23]. Amplification was observed in all tested samples, including the recent variants-of-interest JN.2, XBB, and EG.5 that were identified in our geographic area. Target failure, which was frequently presented in SARS-CoV-2 RT-PCR diagnostic methods [15, 24], was observed only in 1 case, as shown in Figure 5 (sample #3, a sample from April 2021), where the gene S was not amplified. All other samples showed amplification for all targeted genes. The in silico analysis showed a perfect match of the primers against all current SARS-CoV-2 genomes, with the mention that in the case of JN.1 and BA.2.86 variants, a single nucleotide mismatch was found (the 15th nucleotide (C) of the S forward primer mismatched the (T) on the viral genome). Nevertheless, amplification bands with the expected size were visible on gel electrophoresis, and as presented by many PCR kit manufacturers, the amplification of at least 1 out of 3 target genes is sufficient for a positive result [25]. This proves that our selected primers are efficient and able to detect all currently known variants.

Regardless of the price, qRT-PCR was shown to have the best performance characteristics, the specificity, sensitivity, and accuracy reaching 100%; also, the level of detection for some kits went down to 1 viral copy per reaction [12]. TaqPath COVID-19 RT-PCR IVD kit is one of the best performing kits, with a level of detection of 10 viral copies/reaction and a total runtime of 105 min, giving reliability to the processed samples during our endpoint triplex RT-PCR assay development. Also, the calculated sensitivity and specificity of our assay compared to the TaqPath COVID-19 RT-PCR IVD test were 100%, as presented in Table 4.

Recent reviews presented alternative tests for SARS-CoV-2 detection, such as RT-LAMP based assay that targeted the SARS-CoV-2 S gene, with a low cost, but with 88.89% sensitivity, and even lower than 50% after the 10th day after the onset of the disease, thus not comparable to qRT-PCR. Nevertheless, the specificity of LAMP was almost 100%, comparable with the RT-PCR [26, 27]. Nucleic Acid Sequence-Based Amplification, Specific High-Sensitivity Enzymatic Reporter Unlocking or Clustered Regularly Interspaced Short Palindromic Repeats were also described as methods adapted for SARS-CoV-2 detection, with accuracy compared to the qRT-PCR [27], but the latter is still considered gold-standard method by the CDC [28]. Indeed, the rapid PCR testing or point-of-care PCR analyzers and kits, called next-generation diagnostic systems, such as bioMerieux Biofire FilmArray, Cepheid GeneXpert, Abbott ID NOW, or QuantuMDx Q-POC, can offer a quick and safe diagnostic result for SARS-CoV-2 infection with very good sensitivity and specificity, without the needing of molecular biology knowledge or time investment [29‒31]. The main drawbacks of these alternatives are, first, the high cost (a real challenge for the low-income countries of medical facilities), and second, the fact that each test is performed individually, requiring multiple analyzers to efficiently handle a large number of samples in a short time.

Other SARS-CoV-2 detection methods such as the lateral flow assay have shown 57% clinical sensitivity, 69% accuracy, and 100% specificity for IgM antibodies detection, respectively, and 81% sensitivity, 86% accuracy, and 100% specificity for IgG antibodies [27]. In our opinion, it is risky to use for diagnostic purposes a method with a sensitivity lower than 90%, especially in high-risk hospital wards, but our proposed method might be a cheaper, yet accurate method, to be used for low-income countries, with limited infrastructure.

Generally, the laboratory management aims to generate profit through sample processing, not only in the case of COVID-19 diagnostic, but for all types of laboratory analyses. It is well known that for processing new types of tests it is necessary to invest in new equipment and personal training, or even build new spaces that follow the national regulations. Microbiology departments were initially not equipped for molecular biology, and had to make substantial investments as the COVID-19 pandemic emerged. This included purchasing real-time PCR analyzers, centrifuges, microbiological safety cabinets, and other essential equipment to ensure fast and reliable results.

According to worldometers.info [32], low-income countries have conducted a very small number of qRT-PCR tests for COVID-19 diagnosis. We can only speculate that the limitations in testing could be due to inaccurate reporting of the performed tests, high costs, or limited availability of qRT-PCR. Table 5 shows representative countries found in this situation; for instance, only 0.5% of the Algerian population was tested for SARS-CoV-2, whereas the USA, with a population of 335 million, conducted nearly 1.2 billion tests, marking the highest number of tests performed in any single country [32]. Overall, there is a statistically direct correlation between the gross national income (GNI) per capita [33] and the number of qRT-PCR tests per 1 million population (r = 0.8322, 0.775–0.876 at 95% confidence interval, p < 0.0001), as represented in Figure 6.

Moreover, the number of tests differs significantly in countries with GNI <10,000 and >10,000 USD/capita (median values 300,619 vs. 2,611,693 qRT-PCR tests, p < 0.0001). This supports the hypothesis that low-income countries, due to budget restrictions, failed to perform qRT-PCR tests. For this, cheaper, but equally reliable tests are needed to follow and control COVID-19 pandemics. By comparing the methods described in this study for endpoint RT-PCR, the technical requirements consist of a PCR thermal cycler and a horizontal electrophoresis system, with an investment of approximately 12.000 Euros. On the other hand, for real-time PCR, only the amplification equipment is approximately 60.000 Euros.

As it is presented in Table 6, there are differences between qRT-PCR and endpoint RT-PCR, not only at prices but also regarding the need for different materials. For qRT-PCR, additional materials such as reaction plates or adhesive amplification films contribute to the increased costs. The price for qRT-PCR was calculated for a single analysis, in ideal working conditions (maximum number of samples processed in a 96-well plate); the costs may vary unfavorably for qRT-PCR when processing fewer samples. The cost of qRT-PCR was found to be almost 5 times higher compared to our endpoint RT-PCR. qRT-PCR offers benefits such as reduced processing time, a straightforward methodology, and prevention of potential cross-contaminations from PCR amplification products.

In the case of 2-step RT-PCR, the total amount of time from the sample’s arrival in the laboratory until the release of the results was about 5 h per sample. This can be a disadvantage, but the significantly lower price will compensate. The RNA extraction requires approximately 20 min per sample, followed by 1 h for the reverse transcription, 2.5 h for the amplification, and 1 h for the electrophoresis. Nevertheless, by using 1-step endpoint RT-PCR, the working complexity and time have improved significantly (20 min for the RNA extraction, 1.5 h for amplification, and 1 h for electrophoresis).

At the beginning of the pandemic, immunochromatographic tests were unavailable. Later, due to existing regulations in Romania, these tests were not considered to be effective, likely due to their relatively low sensitivity and specificity. As a result, qRT-PCR testing was widely implemented. The qRT-PCR cost was high compared to rapid tests, possibly due to the lack of an alternative. Many laboratories took advantage of the opportunity and invested in molecular biology equipment. As presented by Domínguez et al. [34], the cost of a qRT-PCR test was estimated at around 45 euro, a price comparable to the one established in our laboratory using the TaqPath COVID-19 RT-PCR IVD kit; nevertheless, other data present qRT-PCR prices for COVID-19 diagnostic between 50 and 100 Euros per test.

The study presents a valid method for detecting SARS-CoV-2 using endpoint RT-PCR, which involves three main steps: RNA extraction, 1-step amplification, and gel electrophoresis. This method can process at least 30 samples in under 3 h, depending on the size of the electrophoresis gel. The method is significantly cheaper than the commercial qRT-PCR methods, which require expensive equipment and reagents, and is able to detect all current SARS-CoV-2 variants, including BA.2.86, JN.1, XBB, and EG.5. Moreover, the presented methodology can be easily adapted for future research and development of other PCR assays in epidemiological need.

We are grateful to Simona Paraschiv, PhD, and Marius Surleac, PhD, for the kind support they provided on sequencing and data analysis.

The study was approved by the Ethical Board of Mures Clinical County Hospital (resolution no. 8665/May 09, 2023), confirming adherence to the principles of ethical standards in human studies as stated by the World Medical Association Declaration of Helsinki. The board also granted the study an exemption from requiring written informed consent due to retrospective use of already stored RNA samples.

The authors declare no conflicts of interest.

The research supported by internal resources, and no external funding was received for this study.

Camelia Vintilă: conceptualization, methodology, resources, investigation, data curation, and writing – original draft preparation. Razvan Lucian Coșeriu: methodology, investigation, and software. Anca Delia Mare: methodology, formal analysis, and writing – review and editing. Cristina Nicoleta Ciurea: methodology, formal analysis, and data curation. Radu Togănel: investigation, data curation, and writing – original draft preparation. Anastasia Simion: investigation, data curation, and visualization. Anca Cighir: resources, formal analysis, and writing – original draft preparation. Adrian Man: conceptualization, methodology, resources, data curation, supervision, and writing – review and editing. All authors have read and agreed to the published version of the manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.