Discussion
We compared aerosol RNA shedding in influenza A cases infected naturally
and by nasal instillation under experimental conditions. Previously, we
cultured influenza virus from exhaled breath showing that quantitative
culture correlates with RNA copy detection (r = 0.34,
p<0.0001).15 A minority of experimental
cases shed virus into aerosols (28%). A far greater proportion of the
naturally infected study population shed into aerosols (86%). Among the
experimentally infected with detectable viral aerosols, the fine RNA
copy GM was within log10 that for naturally infected
cases. A more substantial difference in fine aerosol shedding rate was
observed at the level of the overall distribution, with increases in
median and upper percentiles for naturally infected cases (Figure S3).
Compared with naturally infected cases at the 95thpercentile of fine aerosol shedding, experimentally infected cases shed
nearly 2.5 log10 fewer RNA copies. Given the selection
of naturally infected cases on ILI symptoms and/or a positive rapid
antigen test, and an observed relationship between symptoms and shedding
(Table 3), it is possible that aerosol shedding observed for the
naturally infected study populations overestimates what might be
expected for a sample representative of all naturally-acquired
infections.
Peak RNA copies shed into aerosols on day 3 post nasal inoculation was
one day later than that observed in upper respiratory mucosa in previous
challenge studies,7 yet consistent with the 1-2 day
post symptom onset nose and throat swab viral load peak from household
contact surveillance studies.8 There are no available
data with which to compare the current study’s aerosol shedding. If we
take previous estimates of ~1-2 days as the influenza A
incubation period, plus about 1 day post symptom onset to reach peak
aerosol shedding in the naturally infected population, then we would
expect peak aerosol shedding on ~2-3 days following
exposure to virus. This is consistent with the peak in aerosol shedding
in the experimentally infected cases at 3 days post inoculation,
suggesting that the progression of infection from exposure to
replication is consistent with the natural infection group and with
other studies. The significant decrease in aerosol shedding by day post
onset (Table 3) among the naturally infected, but not the experimental
cases, could be related to failure to detect a clear trend against a
background of very much lower overall aerosol shedding for the
experimental group. The temporal decline observed for naturally infected
cases along with previous reports of less temporal decline in viral load
from nasopharyngeal swabs compared with fine aerosols following day 1
post symptom onset,15 and a tendency for nasal viral
load to overestimate transmission risk after day 3 post symptom
onset,12 suggests aerosol shedding may better fit
epidemiologically observed transmission dynamics over time in the
household setting.
Lower respiratory and cough symptom scores were associated with viral
aerosol detection in both groups with a clear dose response relationship
in cough score for the naturally infected cases, supporting the notion
of a symptomatology-shedding relationship for aerosols. Peak symptom
scores and aerosol shedding coincided, consistent with the temporal
dynamics of other studies.7–9 However, regression
analyses that restricted observations to maximum fine aerosol samples
found mostly weak and unstable effects of symptoms on shedding rate.
Only body temperature in the combined population analysis was
significantly associated with shedding strength. These data should be
interpreted with caution given limited heterogeneity in symptom
severity, with milder illness characteristic of the experimental cases
and more moderate to severe illness characteristic of the naturally
infected cases. It may be for this reason that other studies have
reported mixed results with respect to associations between
symptomatology and nose/throat viral load.9,10 The
current data may nevertheless indicate that febrile periods of illness
are associated with increased infectivity risk. Studying cases with a
broader range of symptom severity could provide additional insight into
symptom-shedding relationships with useful implications for identifying
contagious symptom profiles.
There is growing evidence that airborne transmission plays an important
role in the spread of influenza.5,14,20,21 Humans
experimentally challenged to influenza virus by airborne particles had a
50% risk of infection to a 0.6-3.5 TCID50 dose and
exhibited increased propensity for moderate to severe illness with fever
and cough compared with others experimentally challenged by nasal
droplets.1,2 The term anisotropic has been used to
describe such infections where inoculation mode determines illness
presentation.4,5 A population of cases with
naturally-acquired infections would be expected to demonstrate a higher
proportion of moderate-severe influenza-like illness compared with a
population of cases exclusively infected by exposure to the nasal
mucosa. Compared with other symptoms, systemic scores declined more
rapidly following day 1 post symptom onset for natural cases, consistent
with other findings8 and suggestive of the immune
system clearing systemic infection. These findings may hint that natural
cases may be more likely to result in lung and systemic infection
initiated by an airborne dose, whereas experimentally infected cases
with only nasal mucosal exposure were more likely to have few if any
systemic symptoms and illness more localized to the upper respiratory
tract.
Given the selection bias and the contrasted symptom profiles in two
populations and observed associations between symptoms and shedding, we
attempted to adjust for the effect of symptoms to understand the direct
effect of experimental versus natural infection on the viral load in
fine particle aerosols (Figure S1). Ultimately, propensity score
modelling failed to balance the distribution of covariates between
groups and we concluded that the groups were simply too different to
achieve an unbiased estimate of the main effect of group membership
(experimental versus natural) on shedding strength. Assuming minimal
contribution of potential confounders on the pathway between mode of
inoculation and study population membership (i.e., age, sex, host
immunity, virus pathogenicity, dose) we cannot conclude that the
unadjusted differences in symptomatology and shedding are a result of
mode of inoculation, or simply the result of the differences inherent in
recruitment and enrollment procedures and other potentially unobserved
confounders in the absence of a randomized controlled design.
Identifying naturally infected reference groups that represent the true
distribution of symptom severity presents a challenge. Although a
substantial proportion of cases are asymptomatic, symptomatic community
cases are more prone for inclusion in epidemiologic studies upon seeking
medical attention.22 Multi-year sero-surveillance of
large cohorts in the UK shows influenza infections presented
asymptomatically at a rate of 77 per 100
person-seasons.23 A meta-analysis of longitudinal
studies using serological evidence of infection and controlling for
background illness reported a 65%-85% asymptomatic
fraction.24 It is possible that the experimentally
infected cases are not substantially different in symptomatology from a
representative sample of all influenza infections. Accessing a
representative sample might be achieved through intensive household or
dormitory surveillance of contacts of known cases.
The infectious dose for airborne influenza, and the infectious potential
of cases infected by various modes is largely unknown. If the typical
fine aerosol shedding rate from influenza cases is important for driving
airborne transmission, then our findings would indicate that nasal
mucosal exposure in the experimental challenge model produces cases with
airborne infectious potential similar to symptomatic cases infected
naturally by contact, large droplets, or fine aerosols. If above average
shedders are important for driving airborne transmission (i.e.,
superspreader hypothesis), then infections acquired through nasal mucosa
may not pose as much airborne infectious potential. If we assume
hypothetically that the symptomatic naturally infected cases drawn from
the University of Maryland campus community represent the upper 1% of
symptom severity and shedding strength in a broader population, and if
we also assume that the experimental cases are representative of total
community infections, the chances of an experimental case reaching the
level of fine aerosol shedding observed in the naturally infected group
would be 0.39% (1% of 39 experimental cases). If shedders in the upper
percentiles of shedding rate are responsible for driving transmission,
then it would take many more experimental cases to adequately simulate
transmission events in a human transmission challenge trial model. This
introduces logistical challenges and motivates work to identify, among
naturally infected shedders, characteristics predictive of aerosols (and
mucosal) shedding. In particular, response to infection by different
modes may vary between children and adults, with implications for
subsequent infectivity and population
epidemiology.10,13,25,26 Clinical detection of
infections that may be associated with disease severity and potential
for self-isolation or other behaviors that could modify contagiousness
should be considered in population-level transmission risk assessment.
Given bias introduced by selection of natural infections on symptoms
plus fever or rapid antigen test, the observed correlations of symptoms
and fever with viral shedding into aerosols, and a small N with minimal
covariable distributional overlap precluding appropriate adjustment, we
conclude that the naturally infected population is too different from
the experimentally infected cases to make valid comparisons. Our
observations show that the 52 nasally inoculated experimental cases
produced viral aerosol shedders less frequently than the 83 symptomatic
naturally infected population. When they did shed detectable virus into
aerosols, the experimental cases did so at substantially lower
quantities than the symptomatic naturally infected group. This
difference in aerosol shedding was most pronounced when comparing the
highest percentiles of aerosol shedding for each group. The probability
and quantity of aerosol shedding in unselected natural infection is
unknown. Therefore, these findings encourage efforts to evaluate
shedding from infections observed during contact surveillance without
selection based on symptoms or fever.