Rapid lateral-flow immunochromatographic tests to assess anti N/S IgG seropositivity after BNT162b2 vaccine: A cross-sectional study: Rapid lateral-flow immunochromatographic tests after BNT162b2 vaccine.

Dear Editor,

As reported recently in this Journal, antibodies against SARS-CoV-2 can be detected as early as 7–14 days after natural infection and the antibody titre could persist for more than 6 months. The immune response is elicited against several viral epitopes, yet the nucleocapsid (N) protein and the spike (S) protein (with its subunits S1 – containing the receptor binding domain – and S2 – which mediates viral fusion and entry) were those selected to develop diagnostic methods. Anti-S antibodies have been found to correlate with in vitro neutralization activity. Consequently, the S protein was selected as the target for the development of SARS-CoV-2 vaccines.

Rapid lateral-flow immunochromatographic tests (RLITs) can detect IgG and/or IgM antibodies against SARS-CoV-2 proteins N and/or S in capillary blood, serum, and plasma: this point of care method has already been successfully used in population studies.4, 5, 6, 7

Another application of RLITs could be in the qualitative determination of antibody response after SARS-CoV-2 vaccination. In Italy, the immunization campaign started on 27th December 2020 with the priority given to health care workers (HCWs) vaccinated with BNT162b2 vaccine. BNT162b2 has been demonstrated to elicit a robust anti S antibody response in up to 95% of individuals after the second shot. Although the detection of antibodies in peripheral blood samples is the gold standard, it is expensive and needs expert personnel. These barriers could be overcome by RLITs, but studies assessing the performance of RLITs after BNT162b2 vaccine are lacking and whether they could adequately detect this response is unknown.

The aim of our study was to estimate the qualitative antibody response elicited by BNT162b2 vaccine using different RLITs in a sample of vaccinated HCWs at Luigi Sacco Hospital, Milan, Italy.

In this cross-sectional study, we estimated the antibody response to SARS-CoV-2 antigens (N and S proteins) using three different RLITs in a group of vaccinated HCWs. RLITs were performed between 25th January and 16th February 2021, 7 (±3) days after the second BNT162b2 dose. All the hospital staff was invited to participate on a voluntary basis, and everyone gave written informed consent. A questionnaire was filled to assess gender, age, and previous self-reported SARS-CoV-2 exposure (defined as having had a previous positive nasopharyngeal swab and/or a positive IgG serology). The anti-N protein COVID-19 IgG/IgM rapid test (PRIMA Lab SA, Balerna, Switzerland) and the anti-N and anti-S COVID-19 IgG/IgM rapid test cassette (Zhejiang Orient Gene Biotech Co., InnoLiving, Zhejiang, China) were performed on a single capillary blood sample. The anti-N and anti-S COVID-19Speed IgG/IgM test (BioSpeedia SAS – Institut Pasteur, Paris, France) was later available and simultaneously performed only on a subsample of HCWs. RLITs were read by two investigators (LP and FC). Only the IgG band was considered for the present analysis. RLITs IgG results were categorized as positive, negative or indeterminate (if the IgG band was incomplete). All subjects underwent a concomitant anti-S serological examination on peripheral blood assessed by means of Chemiluminescent immunoassay (CLIA) (LIAISON SARS-CoV-2 trimericS IgG DiaSorin, Saluggia, Vercelli, Italy). An anti S titre ≥ 33.8 Binding Arbitrary Units (BAU)/mL on peripheral blood was considered as positive.

To estimate vaccine response, assuming a response rate ≥95% with a 95% confidence interval and a precision of at least 5%, a minimum of 73 subjects was needed. The study was approved by University of Milan's Ethical Committee.

Of the 160 HCWs included in the analysis, 110 (68.8%) were female and the median age was 41 years (Table 1 ). Twenty-six (16%) reported a previous SARS-CoV-2 exposure. All subjects tested positive on anti S peripheral blood with significantly higher titers observed in subjects previously exposed to SARS-CoV-2 when compared to the unexposed ones [6745 BAU/mL (Inter Quartile Range (IQR) 4452–9960) vs 1995 BAU/mL (IQR 1202–3257), respectively; p<0.001]. The anti-N and anti-S COVID-19 IgG/IgM rapid test cassette RLIT resulted positive in 26/26 (100%) of exposed and 129/134 (96.3%) of unexposed HCWs.

Table 1

Characteristics of the study population and different tests’ results. The presence of a clearly identifiable and complete IgG band was considered positive, the complete absence was considered negative and a partial/incomplete band was considered as indeterminate.

		Overall	Self-reported SARS-CoV-2 exposure before vaccination
			NO	YES	p-value
		n = 160	n = 134 (84%)	n = 26 (16%)
Gender, n (%)	Females	110 (68.8)	92 (68.7)	18 (69.2)	0.999
	Males	50 (31.2)	42 (31.3)	8 (30.8)
Age, median [IQR]		41.00 [32.00, 52.25]	41.00 [33.00, 53.00]	34.00 [28.00, 45.75]	0.028
Positive serological test*, n (%)		160 (100.0)	134 (100.0)	26 (100.0)
Ab Anti-SARS-CoV-2 measured by CLIA (BAU/mL), median [IQR]		2125 [1312, 4250]	1995 [1202, 3257]	6745 [4452, 9960]	<0.001
Rapid lateral-flow immunochromatographic tests
COVID-19 IgG/IgM rapid test (anti-N protein), n (%)	Negative	146 (91.2)	132 (98.5)	14 (53.8)	<0.001
	Positive	10 (6.2)	1 (0.7)	9 (34.6)
	Indeterminate	4 (2.5)	1 (0.7)	3 (11.5)
COVID-19 IgG/IgM rapid test cassette (anti-N and anti-S), n (%)	Negative	2 (1.2)	2 (1.5)	0 (0.0)	0.999
	Positive	155 (96.9)	129 (96.3)	26 (100.0)
	Indeterminate	3 (1.9)	3 (2.2)	0 (0.0)
COVID-19Speed IgG/IgM test (anti-N and anti-S), n (%) (n = 88)	Negative	20 (22.7)	17 (22.7)	3 (23.1)	0.374
	Positive	56 (63.6)	46 (61.3)	10 (76.9)
	Indeterminate	12 (13.6)	12 (16.0)	0 (0.0)

*Cut-off for positivity > or = 33.8 BAU/mL.

List of abbreviations: S, spike; N, nucleocapsid; IQR, Inter Quartile Range; n, number; CLIA, Chemiluminescent immunoassay; BAU, Binding Arbitrary Units.

One-hundred and fifty-five out of 160 and 56/88 subjects tested positive with anti-N and anti-S COVID-19 IgG/IgM rapid test cassette and anti N and anti-S COVID-19Speed IgG/IgM test: assuming CLIA on peripheral blood as the reference, this accounts for a sensitivity of 96.9% [95% Confidence Interval (CI) 92.9%−99%] and 63.6% [95% CI 52.7%−73.6%], respectively.

In our study anti-N and anti-S COVID-19 IgG/IgM rapid test cassette showed a good performance in identifying antibody response (96.9%) after the second dose of BNT162b2 vaccine. Whereas, the anti-N and anti-S COVID-19Speed IgG/IgM test identified only 63.6% of subjects with a positive anti S response after BNT162b2 vaccine.

Only two subjects with no known previous SARS-CoV-2 exposure tested negative with anti-N and anti-S COVID-19 IgG/IgM rapid test cassette. Although both subjects tested positive with CLIA, they showed the lowest antibody titre in the cohort (64 BAU/mL and 253 BAU/mL). In addition, both subjects reported autoimmune disorders: in one case atopic dermatitis treated with janus kinase inhibitor and in the other systemic lupus erythematosus treated with hydroxychloroquine.

The observed higher titers in subjects with a previous SARS-CoV-2 exposure when compared to those unexposed before vaccination was in line with previous observations suggesting that just one single dose of BNT162b2 vaccine could be sufficient to elicit an adequate antibody titre.

Our study presents some limitations. First, not all subjects had an available anti-N titre to systematically ascertain SARS-CoV-2 exposure before vaccination and consequently asymptomatic infections could not be definitely ruled out. Second, the study population is a convenience sample of HCWs not representative of the vaccinated general population. Third, a single serological determination was performed, thus not allowing a longitudinal assessment of test performance overtime. In the end, the absence of vaccine non-responders (with a negative antibody titre) does not allow the assessment of different tests’ specificity.

In conclusion, RLITs could be considered for a qualitative assessment of BNT162b2 vaccine antibody response. RLITs could serve as a tool for a rapid point of care evaluation in people at risk of non-response (i.e. those exposed to immunosuppressant agents). In this population a negative result should be further evaluated by means of CLIA.