Science - USA (2022-01-07)

The estimated cumulative incidence of COVID-
19 by the end of blinded follow-up (100 days
after day 57) for the entire vaccine group was
0.0033 (95% CI, 0.0022 to 0.0045). On the

basis of nonparametric threshold regression,

this cumulative incidence decreased across

vaccinated subgroups with day 57 ID 50 titer

above a given threshold, with zero COVID-19

end points at ID 50 titer above 1000 IU 50 /ml

(Fig. 4A). The shape of cumulative incidence

across threshold subgroups tracked the re-

verse cumulative distribution function of ID 50

SCIENCEscience.org 7 JANUARY 2022¥VOL 375 ISSUE 6576 45

Probability of COVID-19 Through

100 Days Post Day 57 Visit

Days Since Day 57 Visit

Binding Antibody to Spike: Day 57

No. At-Risk*

Low:

Med:

High:

4573 4552 4187 1411

4804 4741 4359 1843

4687 4656 4326 1670

Cumulative No. of COVID-19 Endpoints**

Low:

Med:

High:

0 8 14 25

0 7 10 14

035 8

A

C

(< 2190 BAU/ml)

(2190 to 3800 BAU/ml)

(> 3800 BAU/ml)

Baseline covariates adjusted for: baseline risk score, At Risk status, Community of color status.

Maximum failure event time 100 days post Day 57 visit.

¶ Tertiles:

Spike IgG: Low is < 2190 BAU/ml, Medium is 2190 to 3800 BAU/ml, High is > 3800 BAU/ml.

RBD IgG: Low is < 3310 BAU/ml, Medium is 3310 to 5750 BAU/ml, High is > 5750 BAU/ml.

ID50: Low is < 178 IU50/ml, Medium is 178 to 363 IU50/ml, High is > 363 IU50/ml.

ID80: Low is < 407 IU80/ml, Medium is 407 to 661 IU80/ml, High is > 661 IU80/ml.

§ No. cases = estimated number of this cohort with an observed COVID-19 endpoint. The total count (47) across all tertiles for e

from 36 (Figure 1, Table 1), because the 47 includes all vaccine breakthrough cases including the 11 without Day 1, 29, 57 antibody marker data.

No. at-risk = estimated number in the population for analysis: baseline negative per-protocol vaccine recipients not experiencing the COVID-19

endpoint through 6 days post Day 57 visit. No. cases = estimated number of this cohort with an observed COVID-19 endpoint.

FDR (false discovery rate)-adjusted p-values and FWER (family-wise error rate)-adjusted p-values are computed over the set of p-values both for

quantitative markers and categorical markers (Low, Medium, High) using the Westfall and Young permutation method (10,000 replicates).

*No. At-Risk = estimated number in the population for analysis: baseline negative per-protocol vaccine recipients not experiencing the COVID-19 endpoint through 6 days post Day 57 visit.

**Cumulative No. of COVID-19 Endpoints = estimated cumulative number of this cohort with a COVID-19 endpoint.

Low

Medium

High

Low

Medium

High

Low

Medium

High

Low

Medium

High

¦

Plac: 0 187 409 646

Plac: 13758 13218 11165 3364

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0 30 60 100 0 30 60 100

Days Since Day 57 Visit

50% Inhibitory Dilution Pseudovirus

Neutralization Titer: Day 57

Vaccine low

Vaccine medium

Vaccine high

Placebo

Low:

Med:

High:

4727 4705 4384 1391

4681 4635 4260 1669

4656 4609 4228 1865

Low:

Med:

High:

0 10 13 21

0 4 10 18

03 7 8

B

(< 178 IU50/ml)

(178 to 363 IU50/ml)

(> 363 IU50/ml)

No. At-Risk*

Cumulative No. of COVID-19 Endpoints**

Plac: 0 187 409 646

Plac: 13758 13218 11165 3364

Probability of COVID-19 Through

100 Days Post Day 57 Visit

COVE

Immunologic Marker

No. Cases/

No. At-Risk

Hazard Ratio

Pt. Est.

Attack

rate

Tertile P-Value

(2-Sided)

Overall

P-value

FDR-

adjusted

p-value

FWER-

adjusted

95% CI p-value

Vaccine low

Vaccine medium

Vaccine high

Placebo

Fig. 2. COVID-19 risk by antibody marker level.The plots and table show covariate-adjusted cumulative incidence of COVID-19 by low, medium, and high tertiles of
day 57 IgG concentration or pseudovirus neutralization titer in baseline SARS-CoV-2Ðnegative per-protocol participants. (A) Anti-spike IgG concentration. (B) ID 50
titer. (C) IgG (spike and RBD) and pseudovirus neutralization titer (ID 50 and ID 80 ). The overallPvalue is from a generalized Wald test of whether the hazard rate of
COVID-19 differed across the low, medium, and high subgroups. Baseline covariates are adjusted for baseline risk score, at risk status, and community of color status.
Pt. Est., point estimate; FDR, false discovery rate; FWER, family-wise error rate.

RESEARCH | RESEARCH ARTICLES