To the Editor:

In December 2019, China notified the World Health Organization (WHO) regarding the several cases of Severe Acute Respiratory Syndrome (SARS), later named to be COVID‐19, caused by a coronavirus named SARS‐CoV‐2. This pandemic has 6 continents under its grip and the attempts at flattening the curve have become more crucial and frivolous. One such promising modality is the development of a vaccine against the same with Oxford vaccine (ChAdOx1 nCoV‐19), based on a chimpanzee adenovirus, being at the forefront.¹

One of the fears that grapples the success of these vaccines is the phenomenon of antibody‐dependent enhancement (ADE) or more historically vivid‐ Trojan Horse phenomenon. The concept of ADE was given in 1964 by Hawkes et al who in vitro experiments demonstrated increased infectivity of Murray Valley encephalitis virus in the presence of antibodies.² Consequently, these findings were extrapolated to Dengue virus infection wherein, non‐neutralizing antiviral antibodies facilitate virus entry into host cells, leading to enhanced infectivity. Instead of neutralizing Dengue virus, these antibodies bind to the viral proteins in such a way that actually helps them to invade the immune system and further spread.^{3, 4}

Another subtype of heterotypic antibodies is produced which act as the infamous Trojan horse of Troy.³ These antibodies are non‐neutralizing and are delivered to antigen‐presenting cells (APCs) where they undergo aberrative processing after phagocytosis leading to high virion production, while surpassing the immunosurveillance. This may cause severe and potentially life‐threatening syndromes like dengue haemorrhagic fever and/or dengue shock syndrome, in case of reinfection.

Both, ADE and Trojan horse antibody phenomenon can be largely attributed to restricted viral gene expression and non‐binding of antibodies to the right region of the virus. Due to low levels of virus gene products in the infected cells, they escape detection and destruction by host immune surveillances. However, tissue damage occurs due to the inflammatory response generated against viral antigens.³

Similar phenomenon has been reported in Japanese encephalitis, West Nile fever, Respiratory syncytial virus, Zika virus, Influenza virus (particularly pH1N1 outbreak) and lastly, coronaviral severe acute respiratory syndrome (SARS) and Middle East respiratory syndrome (MERS) animal models.² This phenomenon may also explain the severe damage caused to the pneumocytes in COVID‐19, MERS and SARS infection, culminating in acute respiratory distress (ARDs).² Initially, Kam et al investigated the mechanisms underlying SARS‐CoV mediated ADE, where they reported protective immune response in vivo with immunization with recombinant full‐length SARS‐CoV spike S‐protein, but in vitro viral infection of human B cells.² Wang et al noted that antibodies against SARS‐CoV spike proteins rather than nucleoproteins are responsible for ADE.² Many studies did not find any correlation between the clinical outcome and anti‐SARS‐CoV antibodies,² while few studies noted inverse correlation between early SARS‐CoV seroconversion and prognosis of the patient.²

Herein, we report two patients who initially tested positive for SARS‐CoV‐2 and after clinical and virological cure developed a severe complications culminating in ARDs (Table 1). Both the patients achieved the status of de‐isolation and discharge put forth by Centre of Disease Control (CDC), USA.⁵ Interestingly, during readmission there was a rise in viral titres as well as a rise in lactate dehydrogenase (LDH) and C reactive protein (CRP) which heralded the onset of cytokine storm in the patients.⁶

We speculate that ADE mechanism might have played a role in development of fulminant ARDs in our patients. Researchers from Wuhan have also noted reinfection in patients, who were completely recovered from COVID‐19.⁷ It can be argued that this is a flare up of disease rather than a reinfection, however, rising viral titres refutes this idea. Another concern is that residual RNA material from the lysed virion which might be responsible for a positive result on RT‐PCR, as RT‐PCR does not differentiate between live or lysed viral particles.⁸

By extrapolating these findings, we would also like to raise concerns regarding the development of vaccines based on attenuated viruses capable of producing such ‘trojan’ antibodies. This issue was reported earlier in small subset of SARS‐CoV vaccine studies in which vaccine provided protection against SARS‐CoV infection, but resulted in Th2 directed pulmonary damage. In another studies, post‐vaccination challenge of mice with SARS‐CoV nucleocapsid protein and double inactivated SARS‐CoV vaccine failed to provide complete protection and caused severe pneumonia due to enhanced eosinophilic pro‐inflammatory pulmonary response.²

Identification of viral epitopes (eg SARS‐CoV spike proteins and not nucleoproteins are responsible for ADE) is important for a successful production of vaccine, as it heralds a precarious balance between ADE or neutralization of infection.

Thus to conclude, treating physicians need to be wary of the fact that reinfection might be a possible outcome in COVID‐19 patients which may lead to a more severe presentations rather than an innocuous symptoms which the patient initially manifest with. Such cases may also underscore an important immunological concept which might serve as an important steering factor in a safer and more efficacious vaccine development.