Diagnosis of Acute Dengue Virus Infection Using Enzyme-Linked Immunosorbent Assay and Real-Time PCR

Dengue fever is a viral infection caused by the dengue virus and is a growing concern for public health worldwide, particularly in tropical and subtropical regions. This study aimed to assess the diagnostic accuracy of a commercially available NS1 ELISA kit for dengue fever in Pakistan using multiplex qRT-PCR as the gold standard. The study recruited 1236 suspected cases of dengue fever admitted to public sector hospitals in Lahore, Pakistan. Of the suspected cases, 610 (49.3%) were confirmed positive for DENV infection through qRT-PCR, with all four serotypes detected. DENV-2 was the most prevalent serotype, detected in 95.7% of cases. The NS1 ELISA kit detected 71.1% of the positive cases. However, the diagnostic accuracy of the NS1 ELISA kit was found to be only 64.89%. Of the 610 confirmed cases, 68% were male and 32% were female, with a median age of 30 years. Dengue fever was diagnosed in 91.8% of cases, while 8.2% were diagnosed with dengue hemorrhagic fever (DHF). DHF patients had a higher prevalence of abdominal pain, hemorrhagic manifestations, and thrombocytopenia. The cocirculation of all four DENV serotypes in Lahore is concerning and could lead to more severe forms of the disease, such as DHF or dengue shock syndrome, in the future. The study highlights the low diagnostic accuracy of commercially available NS1 ELISA kits and emphasizes the importance of using molecular methods to confirm acute dengue infections. Given the increasing prevalence of dengue fever in developing countries like Pakistan, more accurate and reliable diagnostic tools are needed for effective disease management and control.


Introduction
Globally, the incidence of dengue fever has increased dramatically in recent decades and has become a public health threat of grave concern. About 390 million people contract dengue virus (DENV) infection yearly in more than 128 countries, resulting in 40,000 deaths annually [1,2]. Among them, 92 million develop symptoms of variable intensity, ranging from self-limiting dengue fever (DF) to lifethreatening dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS). DENV is a positively sensed nonsegmented RNA Flavivirus belonging to the family Flaviviridae. Te virus is transmitted chiefy by Aedes aegypti (Ae. aegypti) and Ae. albopictus mosquitoes. Te virus has four antigenically distinct serotypes, including DENV-1, DENV-2, DENV-3, and DENV-4, with variable geographical distribution around the globe [3,4]. Primary infection with one of the DENV serotypes confers long-lasting immunity against the respective serotype. In contrast, the secondary infection with the heterologous DENV serotype might result in a more severe form of the disease (DHF/DSS) through a less understood pathway of antibody-dependent enhancement (ADE) [5]. Currently, no truly safe and effective antiviral agent or vaccine is available against the DENV infection. In the absence of approved therapy, fuid replacement therapy, the use of the antipyretic drug acetaminophen, and other symptomatic treatments remained the mainstay of therapy for DENV infection [2].
Although no part of the world is free from dengue, it is more common in the tropical and subtropical regions where Pakistan is also located. In these regions, cases of dengue fever start to emerge rapidly after monsoon rainfall each year [6,7]. Cases of DF were reported for the frst time in Lahore, Pakistan, in 1968 [8], and several outbreaks have been reported afterward [9]. Te frst signifcant outbreak of DHF occurred in 1994 in Karachi, followed by another in 2005. In 2011, dengue spread to various regions of Punjab and caused morbidity and mortality on a larger scale. Now DF has become endemic, with an annual recurrence in diferent cities in Pakistan [10].
Early and accurate diagnosis of dengue is the primary clinical goal for evidence-based management, vector surveillance, and control measures. Currently, commercially available diagnostic techniques mainly include the detection of DENV antigens or antibodies through rapid immunochromatography (ICT) devices, enzyme-linked immunosorbent assay (ELISA), and detection of viral nucleic acid through reverse transcriptase polymerase chain reaction (RT-PCR) [11]. During the acute phase of dengue fever (up to 6 days after the onset of illness), isolation of the DENV and nucleic acid detection are the ideal methods for diagnosing infection [12]. Viral nucleic acid detection is much easier and requires less expertise, competence, and infrastructure than viral isolation. Viral nucleic acid detection through RT-PCR is more sensitive and specifc, with a faster turnaround than viral isolation methods. Te development of multiplex real-time RT-PCR further enabled the detection of all four DENV serotypes in a single reaction tube, thus making it a highly suitable technique for rapid diagnosis of DENV infection and large-scale surveillance studies [13].
On the other hand, viral antigen detection methods rely on the detection of nonstructural protein-1 (NS1) antigens in the patient's sera. Te NS1 antigen is a glycoprotein required for the replication and viability of DENV. Te NS1 antigen is secreted by mammalian cells in high concentrations, becomes detectable as early as day 1 of the onset of illness, and remains detectable for up to 5-6 days [13]. Terefore, it is valid for the early detection of DENV infection in both primary and secondary infections. Many commercial ELISA kits are available to detect DENV NS-1 antigen with high sensitivity and specifcity, but these kits do not diferentiate among various DENV serotypes. Te current study compared the diagnostic accuracy of commercial ELISA and qRT-PCR kits to diagnose DENV infection in the patient's sera.

Ethical Considerations.
Ethical clearance for this study was taken from Institutional Review Committee, Punjab AIDS control program (PACP-IRC). Written informed consent was taken from all the study participants or their legal guardians in the case of minors.

Study Site and
Sampling. Tis prospective laboratory study was conducted during peak dengue season over a period of 6 months from August 2021 to January 2022 (n � 1236) in the Punjab Public Health Reference Laboratory (PPHRL) and Punjab Aids Control Program (PACP) Associated with the Primary and Secondary Health Care Government of Punjab, Lahore. Te samples were collected from tertiary care hospitals, including Mayo Hospital and Sir Ganga Ram Hospital, Lahore, Pakistan.
Clinically suspected cases of dengue presenting with fever from 1 to 9 days, along with three more symptoms of dengue fever, including headache, myalgia/arthralgia, retroorbital pain, rash, abdominal pain, and hemorrhagic manifestations, were recruited in this study. In contrast, patients who presented with a fever for >10 days and had a fever with the associated focal source of infection (e.g., otitis media, pneumonia, and meningitis), chronic illness, including anemia, and unstable vital signs were excluded from the study. Dengue cases were classifed into dengue fever (DF) and dengue hemorrhagic fever (DHF) according to the World Health Organization's (WHO) classifcation and case defnition [14].
A 3-5 mL blood sample was collected from each participant under aseptic conditions. Te sample was divided into an EDTA vial for a complete blood picture and a gel tube for serum separation. Te demographic and clinical characteristics of each participant were noted on the data collection tool. Te samples were transferred to the laboratory for further testing. Te serum was separated by centrifugation at 5000 rpm for 10 minutes and stored at −80°C for further processing.

Complete Blood
Picture. Whole blood collected in an EDTA vial was used to determine a complete blood picture using an automated hematology analyzer (Sysmex KX-21, Japan). Hematology low (Eightcheck ® 3WP-L, Sysmex) and normal (Eightcheck ® 3WP-N, Sysmex) controls were used to check the performance and calibration of the hematology analyzer. Reference intervals for the interpretation of CBC results were adopted from the previously published criteria [15]. A leukocyte count of less than 4500/mm 3 was defned as leucopenia. A platelet count of less than 150,000/mm 3 of blood was defned as thrombocytopenia. A hematocrit (HCT) of more than 54 and 47 mL/dL in males and females was considered elevated. A hematocrit of more than 49 mL/ dL was taken as elevated in children.

DENV NS1
Antigen Detection Using ELISA. NS1 antigen ELISA was performed on all sera samples using a commercially available Dengue Ag RecombiLISA kit (CTK Biotech USA, catalog #E0312). Te kit has a sensitivity of 100% and a specifcity of 97.8%. Te manufacturer's instructions were followed for performing the ELISA, and each batch was evaluated using positive and negative controls in duplication and interpreting the results as reactive, nonreactive, and equivocal. Washings were performed using an ELISA microtiter plate washer (Combiwash, Human Gmbh, Ref. No. 18460). Te optical density (OD) was measured at Gmbh, Ref. No. 16670). A cut-of greater than 0.3 indicated a positive sample, and a cut-of less than 0.3 indicated a negative sample.

RNA Extraction and Real-Time PCR (qPCR)
. DENV RNA was extracted from serum samples using a commercial nucleic acid extraction kit (Zybio, China) using the manufacturer's protocol. After extraction, DENV RNA was amplifed using a real-time reverse transcription PCR kit (Bosphore ® Dengue virus Genotyping kit v1, Catalog # ABDEG3) to detect all four DENV serotypes in a single reaction tube. Te PCR kit was used with the QuantStudio real-time PCR system (Termo Fischer Scientifc USA). Te kit has a high analytical sensitivity to detect as low as 500 copies/mL of the sample. Te kit was stored at −20°C before use, and components were placed on a cooling block during preparation for PCR. Internal and positive controls were used to determine the assay's performance.

Statistical Analysis.
Te data were entered and analyzed in Microsoft Excel and Statistical Package for Social Sciences (SPSS). A histogram was plotted to determine the distribution of the participants into diferent age groups. Sensitivity, specifcity, positive predictive value (PPV), negative predictive values (NPVs), and concordance were calculated to determine the test's diagnostic accuracy.

Results
A total of 1236 suspected dengue cases reported to public sector hospitals in Lahore were recruited in this study. Te study population included 820 (66.3%) males and 416 (33.6%) females. Te median age of the participants was 30 years (range � 1 to 90 years). Tere were 106 (8.6%) children (≤16 years old) and 1130 (91.4%) adults.
A total of 560/610 (91.8%) cases were diagnosed with DF, while 50 (8.2%) were diagnosed with DHF. Te patients were reported with diferent signs and symptoms, including fever in 610 (100%) patients, followed by abdominal pain in 435 (71.3%), retro-orbital pain in 436 (71.5%), headache in 362 (59.3%), and hemorrhage in 252 (41.3%) patients. Te clinical presentation was also compared according to disease severity. Abdominal pain and hemorrhages were signifcantly associated with DHF (p value <0.05). Similarly, severe thrombocytopenia was reported more commonly (p value <0.05) in DHF patients. Hematocrit, the hallmark of plasma leakage, was signifcantly higher in DHF patients (Table 2). No signifcant diference (p value >0.05) was observed among the age groups of the patients according to disease severity.

Discussion
Currently, no specifc antiviral agent is available for treating DENV infection; therefore, early diagnosis remains the mainstay for the timely management of dengue fever to prevent any complications and severity of the disease [16]. In the current study, we compared the diagnostic accuracy of ELISA to detect DENV NS1 antigen using qRT-PCR as a gold standard.

Canadian Journal of Infectious Diseases and Medical Microbiology
Te current study demonstrated that DENV infections were more commonly detected in males (66.34%) than in females (33.65%). It may be due to the high ratio of outdoor working activities among males compared to females. A study in Uttar Pradesh, India, in 2020, by Kumar shows that the proportion of males is more than females (1.54 : 1), which  supports our study [17]. Another study showed that males were more afected than females (1.5 : 1) due to their frequent outdoor activities [18].
Te most afected age group in the recent study was 15 to 30 (41.1%) years old, followed by 31 to 45 (23.8%) years. Similarly, a study conducted in Pakistan in 2010 demonstrated that the age group of 20-to 30-year-olds was more afected by DENV infections than other age groups [19]. A retrospective study performed in China in 2020 showed a high prevalence among 31-to 45-year-old and 15-to 30year-old age group, while a low incidence of dengue was seen in <14-year-old and >60-year-old age groups [20]. Te highest prevalence of DENV infection in this particular age group was demonstrated by the fact that the highest proportion (63%) of the Pakistani population comprises youth aged 15 to 33 years [21].
Socioeconomic status plays a signifcant role in the transmission and distribution of DENV infections in the community [22]. Interestingly, the highest prevalence (16.9%) of DENV infections was reported among construction laborers, highlighting the public health signifcance of surveillance and insecticide spraying of buildings under construction and buildings not owned by the general public (like Government buildings). Construction sites are the ideal breeding sites for mosquitoes as water is stored in open plastic containers for longer durations, water pockets form during construction sites, there is higher relative humidity, and low light penetration is possible inside the buildings under construction. Moreover, excessive sweating during work also makes laborers more attractive to mosquitoes.
Te prevalence of high DENV infections among students (13.1%) and teachers (9.2%) also identifes the failure of integrated vector management activities at public sector institutes.
Astonishingly, healthcare workers were the third-most afected group (11.8%) after the construction laborers and students. Healthcare professionals usually acquire DENV as a nosocomial infection during patient care and treatment, particularly in endemic areas [23]. One of the reasons could be their vicinity to the dengue patients, thus increasing the chances of human-to-human transmission during a mosquito bite. It also identifes the loopholes in the vector control measures at public sector hospitals and the carelessness on the part of the healthcare workers regarding the preventive measures to avoid mosquito bites.
Garages and motor vehicles are the other neglected sites where mosquitos might be found. Vehicles do not provide ideal breeding sites, but they serve as excellent blood-feeding sites as people (including children) often sit close to each other and are inactive. Among all others, drivers spend most of their time inside these vehicles and might get exposed to mosquito bites for the most extended duration. A sufcient number of dengue patients were drivers (4.1%) in this study.
Te current study detected all four DENV serotypes circulating in Lahore, Pakistan, which is alarming as it might result in an outbreak of more severe DF. DENV-2 was the most prevalent (95.4%) serotype detected in dengue patients. Many extrinsic (viral) and intrinsic (host/physiological) factors play an essential role in the outcome of DENV infection [24]. DENV-2 was associated with severe outcomes of DF [24]. However, this study found no signifcant association between DENV serotypes and severe outcomes (DHF). Infections with multiple DENV serotypes are associated with the severe outcome of DF [25,26]. However, no signifcant association with concurrent infection was noted in this study.
Te current study also evaluated the diagnostic accuracy of ELISA NS1, taking qRT-PCR as a gold standard. ELISA NS1 is more widely available in Pakistan, while qRT-PCR is available only in specialized facilities. Terefore, it is signifcant to determine the diagnostic accuracy of the ELISA NS1 kit against a gold standard. NS1 antigen is a highly conserved glycoprotein in Flaviviruses that can be detected in the patient's serum as early as day one and usually remains detectable till day 9 of the onset of the symptoms. Tus, it serves as an important marker for the early diagnosis of DENV infection [27,28]. ELISA is a powerful technique and is widely available worldwide due to its cost-efectiveness, reliability, and requirement of less sophisticated facilities compared to qRT-PCR. In contrast, qRT-PCR is rapid, much more sensitive, and specifc than ELISA [29]. Te current study determined the low diagnostic accuracy of DENV NS1 ELISA (64.89%) by taking quantitative multiplex qRT-PCR as a gold standard. All the samples suspected of acute DF must also be tested by qRT-PCR for fnal confrmation. Another vital factor that could result in the degradation of the DENV RNA is unsuitable storage and transportation conditions, thus falsely resulting in a decline in the number of positive samples through qRT-PCR [30], thus afecting the diagnostic accuracy of the kit.

Conclusion
After analyzing the data, this study has identifed a concerning situation of cocirculation of all four DENV serotypes in Lahore, with DENV-2 being the most common. Such a scenario could lead to a potential outbreak of a severe form of the disease. Tus, early detection of dengue fever is crucial to prevent complications, economic loss, and mortality. Tis study evaluated the diagnostic accuracy of the NS1 ELISA kit, revealing a sensitivity of 86.72% and a specifcity of 43.61%. Additionally, our fndings indicate that patients with severe thrombocytopenia and higher hematocrit levels are more susceptible to developing DHF. Notably, the study highlights the role of occupation in the spread of vector-borne diseases in communities, with construction laborers and healthcare workers being identifed as high-risk groups for dengue virus infection in Lahore. Overall, these fndings provide valuable insights for policymakers to develop targeted interventions to control the spread of dengue virus infection in Lahore, ultimately leading to improved public health outcomes.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request.

Ethical Approval
Te study was permitted by the Institutional Review Committee, Punjab AIDS control program, Lahore, Pakistan.