RESULTS:
This study was conducted at CMC Vellore with recruitment between
November 2020 and August 2021. The sample size at 80% power and
two-sided 5% significance to test a sample correlation coefficient
(0.54) of IgG levels and time post-symptom onset against a population
correlation coefficient of 0.72 required recruitment of 92 children.
Total of 103 children were recruited, but 24 were excluded (loss to
follow-up or missing data or insufficient blood for antibody testing).
Samples for antibody testing from the cohort (N=79), were collected
between December 2020-August 2021. Based on sequencing data from CMC
Vellore, predominantly wild type (Wu Hu-1/D614G) or delta (B.1.617.2)
were detected from June 2020-March 2021 and April 2021- August 2021 (end
of the study period), respectively.18
Out of the total of 79 children recruited, 29% (23/79) were
asymptomatic and incidentally detected as PCR positive and 59% (47/79)
presented with mild-to-moderate disease not requiring intensive care.
The remaining 11% (9/79) manifested severe disease at presentation
including 2 children with a diagnosis with MIS-C. There was no mortality
in our cohort. The month-wise distribution of date of respiratory and
blood sample collection of study subjects is shown in Figure 1. Sixteen
(20%) of the 79 children tested PCR positive during the wild-type wave
and 59 (75%) during the Delta wave and corresponding sera collected
within the period of circulation were designated “wild-type infected”
and “delta infected”, respectively. Four samples PCR positive during
the wild-type wave, had blood collected during the delta wave, were
classified as ‘SARS CoV-2 infected’. We compared between variants among
the “wild-type infected” (N=16) and “delta infected” (N=59) sera.
Baseline characteristics of children recruited in the study are
presented in Table 1. The median (IQR) age of recruited subjects was
9(11) years. Wild type infection resulted in severe disease in 25% of
children while majority of Delta infections (69%) were
mild-to-moderate. The mean (SD) Ct value were 33.6 (3.4)
and 26.3 (6.7) for wild-type and delta infection, respectively
(P<0.0001). Wild-type infected sera showed 100% positivity to
both anti-N and anti-S antibodies, while delta-infected sera showed 93
and 95% seropositivity, respectively. Anti-S levels were observed to be
higher than anti-N levels (p<0.0001) in both wild-type and
delta-infected sera.
We assessed, if difference in antibody levels observed, were due to
viral load in the respiratory samples collected 4-6 weeks prior to blood
collection. We stratified detected Ct value into
quartiles - Q1(range 10.8-24.8), Q2(range 24.9-27.8), Q3 (range
27.9-32.7) and Q4 (range 32.8-42.1) and compared antibody levels across
them. We found anti-S levels to be higher within each Ct quartile
(p<0.001). When comparing levels between quartiles, we found
that for both anti-N and anti-S antibody, Q1 levels were higher than Q3
(p=0.02 and 0.03, respectively) (Figure 2A). When comparing levels by
variant, we found delta-infected sera (Figure 2B) to show significant
difference in anti-S levels between Q1 and Q3 Ct quartiles . We also
noted higher anti-S levels at Q3 loads between wild-type and
delta-infected sera (Figure 2B). These results suggest a viral load
driver of antibody response and a differential antibody reactivity
between wild-type and delta variants.
We further examined if the observed difference in Ct value between
wild-type and delta varied by clinical severity at presentation. The
mean(SD) Ct of SARS CoV-2 in respiratory samples among
asymptomatic and mild-to-moderate disease was 30.7(4.8) and 26.3(6.8)
(p=0.03), respectively, while among severe disease was 29.7(8.5).
Comparing between variants, we found higher mean (SD) Ctvalues of wild-type [32.5 (4.5)] than Delta [25.0(6.4)] in
mild-to-moderate disease (ANOVA; pairwise comparison p=0.0083) (Figure
3). No difference in wild-type Ct was seen across
severity strata (Figure 3). These results indicate that for delta
infections a higher viral load (lower Ct) was a driver of illness
severity in mild-to-moderate disease. Notably, the Ctvalue range for delta in respiratory samples in our cohort was 10.7-37.7
[~log10(9) difference] among
mild-to-moderate disease (Figure 3).
We further explored antibody levels across severity strata and found
higher anti-S levels compared to anti-N among asymptomatic [median
(IQR) 119.4 (200) and 18.6(66.4), p=0.02] and mild-to-moderate disease
[176.6(147.8) and 50.2(68.9)], p<0.0001], but similar
levels of anti-N and anti-S among severe disease (Figure 4A). When
comparing levels by variant, we found only delta-infected sera to show
differential levels of anti-N and anti-S only for mild-to-moderate
disease [160 (155.1) versus 59.1 (71.0) , p=0.003], but not other
strata (Figure 4B). In the severe disease strata, higher levels of
anti-S among wild type sera compared to delta sera were seen (P=0.02).
These suggest a skewing of antibody response towards S in asymptomatic
and mild-to-moderate infections as well as a differential anti-S
response between variants.