Evaluation of four commercial real-time RT-PCR kits for the detection of dengue viruses in clinical samples

Dengue is caused by 4 related viruses, (DENV)-1, -2, -3 and -4, and is the most common and widespread arthropod-borne viral disease in the world. This self-limited infection can be asymptomatic or cause a clinical spectrum ranging from mild fever (DF) to severe, life-threatening illness, dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS). Other severe manifestations, including hepatitis, myocarditis or encephalitis, may occur independently of plasma leakage [1]. Because dengue symptoms are similar to those of many other undifferentiated febrile syndromes, new commercial tests that can be used for case management have to be clinically evaluated. Laboratory-based surveillance is also of importance for early warning of dengue outbreaks and optimization of vector-control measures [2].

Current diagnostic methods have limitations. Enzyme-linked immunosorbent assays to detect immunoglobulin M are used by most laboratories, but that response is absent early during the disease course and may remain undetectable in secondary dengue. DENV-NS1 antigen can be detected during the first few days of fever. However, recent evaluations of commercial NS1 assays have had limited sensitivity [3, 4]. Virus isolation from acute-phase sera is useless for patient management, because several days are needed to obtain final results. Molecular methods yield same- or next-day detection of DENV in acute-phase serum or plasma. The 2-step hemi-nested reverse transcriptase-polymerase chain reaction (RT-PCR) protocol, originally reported by Lanciotti et al. [5] and later modified to a single-step multiplex RT-PCR [6], has proved to be highly sensitive [7], and is used worldwide. Conventional RT-PCR is being progressively replaced by real-time RT-PCR, which limits the risks of carryover contamination and is suitable for DENV-genome detection. Another advantage of real-time RT-PCR is that it yields reliable quantification, making it a useful tool for pathophysiological studies. Almost all of the reported RT-PCR methods are “in-house” procedures [8‐29]. International external quality studies have highlighted sensitivity and specificity heterogeneities, even inter-laboratory using the same method [30].

Levi et al. [31] evaluated a commercial real-time RT-PCR kit (RealArt; artus/Qiagen, Germany) and found high agreement with an in-house multiplex RT-PCR, but the test was limited to DENV-3, and occasionally DENV-2, samples. More recently, the US Food and Drug Administration approved a Centers for Disease Control DENV-1–4 real-time RT-PCR assay [32], but it has been shown to be less sensitive than a laboratory-developed assay, particularly for DENV-1 [33]. Herein, we evaluated the performances of 4 recently developed commercial DENV real-time RT-PCR kits.

The study was conducted in the Virology Laboratory, where biological samples from patients with acute-febrile syndromes are processed as follows: a 5 ml venous blood sample, collected on the day of admission, is centrifuged, and plasma or serum supernatant is divided into 2 aliquots, one stored at -4°C for daily RT-PCR and the other frozen at -80°C until use. We analyzed 162 laboratory-confirmed stored samples collected during the outbreaks that occurred the last decade: 46 DENV-1, 37 DENV-2, 33 DENV-3, and 46 DENV-4. A panel of 70 negative specimens was used to determine specificity. In addition, we prospectively tested 111 samples from patients consecutively seen at the emergency unit for acute-febrile syndromes at the peak of the 2010–2011 epidemic.

Because the study was non-interventional, e.g. with no additional samplings or specific procedures for subjects, no consent form from the patient was needed, as stated by the French Public Health code. The consecutive samplings were part of routine diagnosis. The dengue sample collection has been declared under the number DC-2009-979 to the French ministry of health and the “Comité de protection des personnes Bordeaux sud-ouest/Outre-mer”, according to the Bioethics laws (décret n°2007-1220 du 10 août 2007 relatif aux activités de prélèvement et conservation à des fins scientifiques de collections d’échantillons biologiques). The dengue collection is stored at the “Centre de Ressources Biologiques de la Martinique” (CeRBiM). The CeRBiM database has been agreed by the CNIL (“Commission nationale de l’informatique et des libertés”).

RNA was extracted from 200 μL of stored aliquots using NuCliSens EasyMag (bioMérieux, Craponne, France). All samples were screened by hemi-nested RT-PCR with the dengue virus consensus primers D1 and D2 and the serotype-specific primers TS1, TS2, TS3, and TS4 [5]. Lanciotti’s protocol was adapted using the SuperScript One-Step RT-PCR System with Platinum Taq DNA Polymerase (Invitrogen, Life Technologies, France) [34]. Briefly, the first-round reaction was performed with primers D1 and D2, 1 μl of SuperScript II RT/Platinum Taq mix (Invitrogen), and 5 μl of the RNA template. The second amplification was used forward primer D1 and downstream primers TS1, TS2, TS3, and TS4, with the TaqPCR Master Mix kit (Qiagen). Both the RT-PCR and nested-PCR products were analyzed by gel electrophoresis on a 2% agarose gel with ethidium bromide for ultraviolet visualization. The expected size of the primary PCR amplicons was 511 bp. Nested-PCR bands of 482, 119, 290, and 389 bp were considered specific for DENV-1, DENV-2, DENV-3, and DENV-4, respectively. The sensitivity of the assay has been previously evaluated to 1 PFU/ml with dengue virus titrated controls, and 1–10 molecules/reaction with dengue virus RNA standards obtained by cloning the capsid-prM region into pCR4-TOPO vector (data not shown). Finally, the procedure was validated through participation in a multicenter quality-control study conducted by the French National Reference Center for Arboviruses (Institut de Recherche Biologique des Armées, France).

Four commercial kits were evaluated versus Lanciotti’s gold-standard protocol. All the commercial kits were fully furnished, and were used according to the manufacturer’s recommendations. The characteristics of the 4 kits tested are summarized in Table 1.