COVID-19: Way Forward With Serosurveillance Without Overemphasizing Neutralizing Antibodies

Serosurveillance of the coronavirus disease 2019 (COVID-19) caused by the severe acute respiratory syndrome-coronavirus 2 (SARS-CoV-2) has been lagging behind. Despite reliable enzyme-linked immunosorbent assay (ELISA) antibody tests being cheaper and most often fast (32,41), there is uncertainty whether prior clearance of the SARS-CoV-2 is truly protective. This has contributed to slower uptake of serosurveillance (42). The scientific community has pointed out some concerns related to the suboptimal performance of certain tests (11), although a selection of tests with both—sensitivity and perhaps more importantly specificity of >99%—is available at this point (18,32,41), and this level of sensitivity and specificity should be sought out.

Unlike in the past when natural immunity held higher status, many in the scientific community currently accept only the presence of specific “neutralizing” antibodies as a surrogate marker of immunological protection against viral infections. However, in the case of SARS-CoV-2, we argue that focusing only on identification of neutralizing antibodies may not be necessary to derive benefits from serosurveillance strategies. Indeed, any evidence of spontaneously clearing the symptoms of COVID-19 may be a reliable surrogate marker of individuals' immune competency regardless of its mechanism (i.e., antibody-dependent cell cytotoxicity and natural killer cells) (38). Clearance of COVID-19 symptoms is closely followed by clearance of all viable SARS-CoV-2 [shown in studies with polymerase chain reaction (PCR) positive/culture negative postinfectious period (43)].

Based on the timing of seroconversion with respect to the duration of viable virus shedding, it is likely that the majority of seropositive patients (particularly IgG positive), at least those with mild/moderate infection, are no longer shedding viable virus (31,33). The inability of currently used PCR testing for SARS-CoV-2 RNA lacks important ability to determine the duration of transmissibility, which limits its utility only as the determinant of illness' etiology. Viral outgrowth testing may not be practical for clinical use, although messenger RNA testing as a surrogate for active viral replication could provide an alternative (43,44).

Evidence-based public health interventions require a relatively high level of certainty around the efficacy (and ultimately real-world effectiveness) of any specific intervention. The acceptable level of certainty varies based on the balance between the risks and benefits in specific diseases, and also on the need to act in a timely manner (pandemic vs. slowly emerging pathogen of sporadic occurrence). The risk of adverse outcomes would be greater with highly lethal acute viruses or with highly virulent chronic viruses for which there are no effective treatment options. SARS-CoV-2 does not belong to either one of those categories and the likelihood of incorrect serosorting of individuals leading to exposure is greater without the knowledge of individuals' serostatus (e.g., in space-limited settings, such as nursing facilities or correctional institutions—previously exposed individuals with mild or asymptomatic prior course of COVID-19 would be a safer recipient community of new arrivals with unclear COVID-19 status as compared with people without prior exposure).

The situation remains complex despite current RNA PCR, which is not uniformly used when needed around the world, and particularly in resource-limited settings where the cost of PCR is more prohibitive as compared with serology that could help determine safer cohorting, including during quarantine time when space is limited and PCR testing cannot be done every day for practical reasons even in resource-rich countries. Further limitations of currently used PCR is that it does not accurately detect the cessation of the infectious period due to its inability to recognize replicable from nonreplicable virus. The temporal relationship between the disappearance of transmissible (live) virus and seroconversion (discussed earlier) would support the fact that asymptomatic seropositive individuals no longer shed transmissible virus.

SARS-CoV-2 is not a highly lethal viral illnesses [e.g., rabies (16)] because most people clear COVID-19 without vaccines (or antiviral drugs). Unlike with rabies, it is less important to focus on just neutralizing antibodies since any sufficient immune response that facilitated clearance without a targeted therapeutic intervention will likely immunologically restrict infection and transmissibility during re-exposures. During re-exposures, preprimed humoral and cellular immunity deals with an inoculum, which is a fraction of the viral load that all convalescent patients previously immunologically eliminated at the height of viral replication.

It is interesting that in COVID-19, we believe that there is enough equipoise to study convalescent plasma infusions of 200–600 cc as a treatment of patients with active replication (8), whereas hesitating on serosurveillance due to fear that convalescents with ∼2.5 L of their own plasma and specifically trained cellular immunity may not be protected from a relatively low inoculum re-exposure.

Viruses with a tendency toward chronicity understandably require greater focus on neutralizing antibodies to overwhelm the odds of establishing chronicity that is quite high in hepatitis C virus (HCV) (75–80%) or HIV (∼100%), and thus the vaccine development success in these diseases has been low (with some exceptions, e.g., hepatitis B virus [HBV]) (13). In comparison with chronic viruses such as HBV, HIV, or even herpes viruses with their unique form of immunological escape (12), in severe acute respiratory viruses, such as SARS-CoV-1, MERS-CoV, and SARS-CoV-2—failed clearance results in death as opposed to established chronicity. Therefore, a survivor of COVID-19 is analogous to individuals who would clear HIV or HBV, and thus defy HIV or HBV chronicity (HBV clearance leaves lasting immunity). Such “lucky” individuals are fortunately much more common with viruses that produce acute disease where dire outcomes develop more rapidly.

An important question pertains to whether convalescent individuals can still become reinfected and contribute to spread. The experience with vaccine preventable diseases and the basic tenets of “herd immunity” (2) would be undermined by such episodes of transmissible virus shedding in immunized individuals. A known example of problems with lasting immunity have been reported for influenza, where a rechallenge with the same strain in just under a year was able to produce evidence of reinfection in a significant proportion of study participants, although the interval between the initial exposure and re-exposure roughly corresponded with expected waning of protection in this setting (26).

Even in chronic viral diseases where vaccine development has so far been elusive (e.g., HCV), reinfected individuals who spontaneously cleared HCV in the past show lower incidence of reinfections and clear them with greater likelihood than individuals achieving sustained virological response to treatment (17). Discussing the impact of various strains of HCV (or similarly dengue virus) on the lack of protective immunity is beyond the scope of this article as it does not pertain to the situation surrounding the single strain SARS-CoV-2, although it is worth mentioning that while antibody-dependent enhancement of infection [similar to dengue (15)] has not been described among other common coronaviruses—this would not necessarily be prevented by vaccine-based strategies (37).

In COVID-19, the vast majority of patients clear all viable/transmissible virus in likely <10 days of infectiveness during the initial episode. This duration is likely even shorter in asymptomatic patients (19). However, the reported range has varied from −6 to +10 days relative to the onset of symptoms (1,4,7,9) and studies evaluating serial cultures/surrogates of activity in their subjects are still sparse (43). The probability that subsequent exposure would lead to clinically meaningful period of transmissibility seems, therefore, relatively low unless such re-exposure occurs when the original immune response wanes (5). To our knowledge, this has so far been shown only in rhesus macaques (6).

Our hypothesis that sufficient postexposure protection can be surmised reliably with SARS-CoV-2 further leans on (1) prior evidence of mostly strain-specific immunity with other coronaviruses (22,23,29); (2) a complete lack of reliable evidence regarding any COVID-19 reinfections so far (20,30) [with a few notable exceptions among >30 million cases worldwide as of mid-September 2020 (39,40)]; and (3) on the fact that natural infection produces lasting immunity, which varies in duration based on the type of virus and its strain (27).

Some argue that more specific and better-controlled immune response may be ultimately achieved through vaccination, and that once successful such approach would be safer than natural immunity (10). This would be particularly important in COVID-19, similar to other viral infections such as dengue virus, where exaggerated immune response has been blamed for certain proportion of adverse outcomes (25). However, the optimal specifics of immunization through vaccination need to be fully elucidated (28) and the pandemic is not forgiving while we learn and test new strategies. In two rhesus macaques studies, natural immunity achieved greater protection upon re-exposure as compared with one type of vaccine despite using a higher viral inoculum for re-exposure in a similar time interval (6,45). Waiting for complete control of the pandemic with sophisticated vaccination or by achieving complete disruption of transmission may take too long.

While waiting for a vaccine development, indiscriminate prolonged lockdown measures causing mass bankruptcy and neglect of other health threats may not be sufficiently justified when we have diagnostic tools (e.g., serology) that can be combined with existing or improved testing methods (e.g., RT-PCR for SARS-CoV-2 RNA or its messenger RNA) (21). Such testing can bring more insight into who needs isolation, and who remains susceptible and why. Immunocompromised individuals and patients with a particularly difficult first episode with prolonged viral replication may choose to continue selective isolation, yet even those would benefit from better understanding of their contacts' serostatus and potential for virus transmission.

The only way to really know is to try serosorting experimentally (34) or to observe such events where they naturally occur (cruise ships, nursing homes, and correctional settings). Furthermore, wide utilization of reliable serological testing could also improve our chance of identifying reinfections in seropositive patients most of whom never had COVID-19 confirmation by PCR due to home isolation measures.

Ultimately, serosurveillance offers a powerful way to relatively cheaply assess prior exposure to COVID-19 at the population level (36). More importantly, it can provide data for evidence-based decision making to restore economic engines, and promote mental and physical health in all settings, particularly in those most vulnerable to adverse outcomes (35). Serological assessments in conglomerate settings such as factories, schools, correctional facilities, nursing/long-term care facilities, and long-distance travel will offer advantages in guiding isolation, quarantine, and cohorting protocols.

For completeness, it is important to mention some of the potential drawbacks regarding serology as a marker of prior exposure. Most importantly, some patients did not develop antibodies despite clearing the virus (14,46), and the duration of detectable antibodies is not clear (24,31,38), although in the largest study today in Iceland the levels of antibodies remained steady for 4 months (through the end of the study). Knowing the duration is important as serology will be a useful tool only if it accurately captures prior exposures. From an epidemiological perspective, however, it is safer to have false negative rather than false positive results.

To comfortably reopen the economy, schools, travel, and so on, we need two things more than anything else: (1) knowledge of how to rapidly identify persons transmitting viable virus and (2) knowing with greater certainty who is and who is not susceptible to COVID-19. The former can be achieved by determining the duration of “live” virus shedding in studies involving viral cultures and by developing rapid testing for replicating virus (e.g., messenger RNA) for clinical use (43). The latter by animal exposure studies, followed by human trials of convalescent re-exposure (3,6), or more slowly, by studies closely following naturally infected patients using vaccine study protocols.