2.1 Misdiagnosis between dengue and Covid‐19

A study consisting of more than 44 000 Covid‐19 cases in China found that 81% of the cases exhibited mild to moderate symptoms.40 Some patients reported complaints of fever, cough, muscle aches, or fatigue with normal chest radiography, resembling other viral infections, thereby making it is difficult for doctors to differentiate Covid‐19 from dengue, which leads to primary examination misdiagnosis.32, 41 Further, clinical manifestations on the skin, such as rashes and petechiae, have been reported in Covid‐19 patients,42 which are also commonly found in dengue.33 A study in Italy found that Covid‐19 patients presented with skin erythematous, rash, urticaria, and vesicles.43 Skin manifestations have also been reported in Madrid.44 Such skin manifestations, including rash or petechiae, increase the challenge in differentiating between dengue and Covid‐19 in dengue‐endemic regions.33, 45 Misdiagnosis between Covid‐19 and dengue based on clinical presentation was also reported in Thailand, where patients presented with petechiae and thrombocytopenia, were diagnosed as dengue but were later diagnosed with Covid‐19 after developing respiratory symptoms and undergoing reverse transcription polymerase chain reaction (RT‐PCR) testing.33, 46 In addition, a study of 41 Covid‐19 patients reported the occurrence of hematological disorders, such as leukopenia, decreased platelet count40 and prothrombin time.45

Beyond the similar clinical manifestations of Covid‐19 and dengue, misdiagnosis may be due to serological cross‐reactivity between SARS‐CoV‐2 and DENV. It was hypothesized that patients with previous exposure to DENV possess anti‐DENV antibodies that are cross‐reactive with SARS‐CoV‐2 antigens.32 Alternatively, there may be antigenic similarities between SARS‐CoV‐2 and DENV, such that upon SARS‐CoV‐2 infection, the body is triggered to generate anti‐DENV antibodies derived from memory B cells.32 These possibilities might have caused the false‐positive phenomena in rapid dengue serology tests.

Misdiagnosis due to serology cross‐reactivity between Covid‐19 and dengue was first reported in Singapore where patients were initially confirmed to have dengue through dengue IgM and IgG rapid serological testing, but were later found to be positive for Covid‐19 using RT‐PCR testing.34 Subsequent RT‐PCR analysis were failed to detect DENV in both urine and blood samples, confirming the absence of dengue infection in the respective patients. This suggests the possibility of serological cross‐reactivity between DENV and SARS‐CoV‐2, in particular when using rapid serology‐based test, that could lead to significant challenges in dengue‐endemic countries in Asia. In Indonesia for example, there have been no reports of misdiagnoses between Covid‐19 and dengue in the literature. However, there has been an increase in dengue cases in the country; 68 000 cases have been reported as of June 21, 2020.29 In Indonesia, most dengue cases were exclusively diagnosed with IgM and IgG rapid serological testing, if not clinically diagnosed.47, 48 Therefore, it is possible that there were Covid‐19 cases among these dengue‐diagnosed patients. In addition, it remains challenging to access Covid‐19 testing in Indonesia, as reported early in the pandemic.49 The lack of adequate laboratories across the country and low testing rate could hamper the diagnosis and management of Covid‐19 patients.

The wide range of Covid‐19 clinical symptoms and lack of a specific and affordable test to differentiate Covid‐19 and dengue have led to serious impacts on community health in the dengue‐endemic countries. The WHO has reported that the antigen‐based rapid diagnostic test (RDT) for Covid‐19 has a sensitivity of 34% to 80%.50 Therefore, there is an urgent need for an affordable and accurate diagnostic method to prevent misdiagnosis of Covid‐19 and dengue in Asian countries with limited resources.

2.2 Shared pathophysiology between Covid‐19 and dengue

Covid‐19 and dengue exhibit some pathophysiological similarities, such as capillary leakage, thrombocytopenia, and coagulopathy (Figure 1).39 Plasma leakage is a crucial factor for dengue pathophysiology and is primarily mediated by the host immunological response. Several immuno‐mediators, including pro‐inflammatory cytokines such as tumor necrosis factor (TNF), interleukin‐6 (IL‐6), interferon gamma (IFN‐γ), and chemokines such as macrophage migration inhibitory factor (MIF), are associated with plasma leakage.51, 52 Altered platelet function53 and increased C‐reactive protein (CRP) levels54, 55 are also responsible for plasma leakage. Covid‐19 is characterized by over‐activation of effector T‐cell function and increased inflammatory cytokine production, especially IL‐6, often leading to a cytokine storm.56 IL‐6, along with other inflammatory mediators, such as IL‐1, TNF, and IFN‐γ, contribute not only to plasma leakage but also to other vascular disorders, including vascular permeability and disseminated intravascular coagulation (DIC).57 In dengue, plasma leakage is associated with the interaction between nonstructural protein 1 (NS1)‐specific antibodies and proteins expressed on endothelial cell surfaces, that may pave the way for the elevated rate of viral replication and inflammatory cytokine secretion.58–60 However, the mechanism of capillary leakage in Covid‐19 is not well understood and further studies are required to elucidate it.

Recent studies have demonstrated that the severity or progression of Covid‐19 is positively associated with plasma CRP levels.61–63 CRP is also important for dengue progression and may potentially be used as a prognostic biomarker.64–67 Hepatocytes primarily synthesize CRP upon stimulation through inflammation. CRP then binds various pathogens and activates the complement system through the classical pathway.68

Thrombocytopenia is one of the main characteristics of DENV infection.69–73 In dengue, NS1 induces complement‐mediated pathways, which cause platelet lysis.74 DENV infected endothelial cells exhibit higher expression of E‐selectin and P‐selectin and increase the interaction of platelets with monocytes, leukocytes, and endothelium, which contribute to thrombocytopenia.75–78 In addition, elevated levels of platelet‐associated immunoglobulin (PAIgM or PAIgG) are reported in dengue cases and are also correlated with thrombocytopenia.74 Thrombocytopenia is also observed in Covid‐19 patients.79–81 A meta‐analysis found that thrombocytopenia is associated with the severity of Covid‐19.79 SARS‐CoV‐2 infection has been speculated to lead to endothelial damage and interfere with the activity of certain receptors to inhibit bone marrow growth and induce apoptosis, leading to abnormal hematopoiesis and ultimately resulting in thrombocytopenia.82 Moreover, immune system‐mediated specific destruction of platelets has also been observed in Covid‐19 patients, which may arise from the elevated levels of immune complexes and autoimmune antibodies.83 It is hypothesized that the cytokine storm damages the bone marrow hematopoietic progenitor cells, resulting in a decline in primary platelet production and a decline in platelet count in peripheral blood.83 However, these speculations must to be investigated and verified in the context of SARS‐CoV‐2 infection.

Coagulopathy is one of the main pathologies associated with DENV infection84, 85 and is also presented in cases of Covid‐19.86–88 Both SARS‐CoV‐2 and dengue infections exhibit prolonged prothrombin time and partial thromboplastin time.89, 90 Surprisingly, the activity of coagulation factors, including prothrombin, factors V, VII, VIII, IX, and X, antithrombin, and α2‐antiplasmin, was markedly reduced during the acute febrile stage of dengue; while in contrast, it appears to increase in the Covid‐19 patients.91 Growing evidence suggests that hypercoagulable state was evident in Covid‐19 patients. This condition leads to the increasing incidence of venous and arterial thromboembolic diseases such as pulmonary embolism, venous thromboembolism, myocardial infarction, stroke, and other microvascular thrombosis (Figure 1).2, 92–96 The pathobiology of coagulopathy in Covid‐19 is not well understood. A recent study reported the presence of antiphospholipid antibodies in Covid‐19 patients, including anticardiolipin IgA antibodies and anti‐β2‐glycoprotein I IgA and IgG antibodies, which abnormally target several phosphoproteins.89 The presence of these antiphospholipid antibodies is considered associated with coagulopathy.97 However, several critical illnesses and other infections also elevated the levels of these antibodies; therefore, it is speculated that the presence of these antibodies might be less associated with the occurrence of any severe coagulopathies.98 Studies have also reported the presence of DIC pre‐terminally in Covid‐19 patients.86–88

2.3 Dengue and Covid‐19: The role of antibody‐dependent enhancement in disease progression

The pathogenesis of Covid‐19 may include antibody‐dependent enhancement (ADE).99 Mechanistically, ADE involves the cross‐linking of virus and antibody or virus and activated complement component complexes through interactions with cellular molecules, such as Fc and complement receptors or cell surface molecules, to promote internalization of the virus and increase the infection of monocytic and granulocytic cells.100–102 Although the role of ADE has been reported in numerous contemporary viral diseases affecting humans and animals,103–105 dengue is arguably the best disease model to study this phenomenon.106 DENV homotypic infection confers long‐lasting and even life‐time immunity. However, cross‐neutralizing antibodies against re‐infection of different serotypes are short‐lived.107, 108 At sub‐neutralizing concentrations, these heterotypic antibodies may trigger ADE and are associated with higher odds of more severe forms of secondary infection when present at relatively low levels in children.109 This condition affects the development of a dengue vaccine, given that vaccination is considered to mimic a primary infection. Indeed, the diminishing levels of antibodies may pose a serious risk of ADE occurrence in children with secondary infections110 and may cause alert for vaccine‐induced enhancement of viral infection.111–113

Approximately 40‐45% of SARS‐CoV‐2 infection may be asymptomatic,114 despite clinical findings explaining the pathological mechanism of Covid‐19 as both directly cytotoxic and immune‐mediated.115 SARS‐CoV‐2 infection induces an exuberant inflammatory response, known as a cytokine storm,116 which subsequently results in uncontrolled pulmonary inflammation primed by rapid viral replication, cellular damage, virus‐induced downregulation of angiotensin converting enzyme 2 (ACE2), viral shedding, and ADE.117 In coronaviruses, ADE may function through the interaction of Fc, or other receptors in susceptible cells, with the complex between virus and anti‐spike (S) protein‐neutralizing antibodies, thereby facilitating both inflammatory responses and persistent viral replication in the lungs.102, 117, 118 ADE have been reported in severe acute respiratory syndrome coronavirus (SARS‐CoV) infection and has been observed in vaccine‐induced infections by a SARS‐CoV vaccine through an Fc gamma receptor II (FcγRII)‐dependent mechanism.119 Polymorphic FcγRIIA on monocytes has been reported to favor ligation of non‐neutralizing antibody‐virus complexes in both dengue and Covid‐19. In dengue, this ligation activates the expression of negative regulators of dihydroxyacetone kinase and autophagy proteins (cascade of down‐regulatory pathways) to decrease production of type I IFN‐activated antiviral molecules and suppress the capacity of macrophage antiviral mechanisms, leading to enhanced viral replication.120 FcγRIIA‐mediated enhancement has been reported in SARS‐CoV in a study of 180 people from Hong Kong who were infected with SARS‐CoV. Moreover, it was found that FcγRIIA polymorphism affects disease severity.121

ADE of SARS‐CoV is thought to utilize a novel cell entry mechanism for entry into immune cells, wherein antibody‐mediated infection is dependent on FcγRII but does not utilize the endosomal/lysosomal pathway through the ACE2 receptor.122 Viral neutralization was demonstrated with concentrated antisera against the S protein, although higher dilutions failed to neutralize the infective virus and even induced apoptosis caused by increased viral infection. In contrast, anti‐nucleocapsid (N) protein antibodies neither neutralized the virus, nor caused ADE. However, nucleocapsid proteins have been observed to induce cytotoxic T lymphocyte (CTL) immune responses that are specifically SARS‐CoV‐2 protective.118

The modulation of pathogenesis among diseases caused by viruses from the same family has been previously described, such as between Zika and dengue.113, 123 Prior infections with other coronaviruses, ranging from those that cause the common cold to SARS, are thought to have primed Covid‐19, and might lead to the development of severe SARS‐CoV‐2 infection.99 However, the molecular and immunological host response to SARS‐CoV‐2 infection have not yet been fully elucidated to confirm ADE.124 In the tropical areas of the world, the apparent cross‐reactivity between antibodies in dengue and Covid‐19 serology tests have been reported.125 A study in Colombia found that there was a decreasing trend of dengue during increasing reports of Covid‐19 and it was speculated it might due to viral interference of SARS‐CoV‐2 over DENV.126 Studies are needed to identify the interaction between these antibodies against dengue and Covid‐19. How will Covid‐19 pathogenesis in populations where the immunities to DENV or other arboviruses are present? Furthermore, the mechanism of ADE in viral infection has become a great concern to disease control by vaccination. Therefore, the development of SARS‐CoV‐2 vaccines has to incorporate approaches to develop the vaccine with minimum or no risk for ADE, if ADE exists in COVID‐19.

2.4 Covid‐19 and dengue cross‐reactivity: Consequences for Covid‐19 vaccine development

Vaccine development must account for safety measures, including immune‐related adverse events (irAE) that may result from vaccination. ADE leading to a cytokine storm is a hallmark of SARS‐CoV pathogenesis and disease severity,127 and of dengue infection.109 This may also be the case for SARS‐CoV‐2.28 ADE has been suggested to be caused by pre‐existing, non‐neutralizing, low‐affinity antibodies,111 along with neutralizing antibody128; therefore, the antigen design of a vaccine must ensure correct antigen conformation129 or spike‐based subunit vaccines lacking the receptor binding domain (RBD).128

A recent study demonstrated that SARS‐CoV‐2 infection provides protection from re‐challenge in rhesus macaques;130 however, the phenomenon in humans requires further investigation. Studies on vaccine candidates should examine whether antibodies raised by vaccine administration result in protection or ADE. Pre‐clinical trials of several SARS‐CoV‐2 vaccine candidates have demonstrated protection after challenge in animals,131, 132 but similar results in humans and the absence of ADE have yet to be observed. Given that anti‐SARS‐CoV‐2 antibodies have been indicated to cross‐react with DENV antigens within dengue rapid serological kits,34 there is a possibility that anti‐SARS‐CoV‐2 antibodies raised by a Covid‐19 vaccine candidate may also cross‐react with DENV antigens. Because ADE in dengue most often results in severe disease progression18, 133 and ADE have been observed previously in other coronavirus,118, 119 it would be advisable to also evaluate the possible effect of ADE‐induced vaccine in Covid‐19.