SARS-CoV-2 and post-donation information: a one-year experience of the French haemovigilance network

Abstract

Background

There is growing evidence to support the hypothesis that SARS-CoV-2 is probably not transmissible by blood transfusion. In this study, we use the data gathered over one year by the French haemovigilance network on post-donation information related to SARS-CoV-2, and virological investigations on corresponding plasma to explore viral transmission by transfusion.

Materials and methods

Whenever a donor reported COVID-19 symptoms and/or a positive SARS-CoV-2 nasopharyngeal (NP) PCR test, information regarding diagnosis and symptoms was collected using a specific questionnaire, and repository plasmas were screened using the SARS-COV-2 R-GENE^® assay (Biomérieux). RNA sequencing (Sanger and deep sequencing) and virus isolation on Vero E6 cells were applied in plasma from donors testing positive.

Results

We investigated 1,092 SARS-CoV-2-related post-donation information (PDI) reports. PDI donors were younger than the global donor population and donated more often in the Paris region. Sixty-eight percent reported a positive NP real-time (RT)-PCR or antigenic testing and 22% of these also had symptoms at the time of testing. Thirty-seven (3.4%) donations tested positive for SARS-CoV-2 RNA, 11 (30%) were confirmed by another molecular assay, and 7 (19%) by sequencing, confirming low viral level. Most RNAemic blood donors donated in southern regions and in Paris. There was no difference in demographic data or duration parameter between RNAemic and non-RNAemic donors. Duration parameter was determined as the time elapsed between donation and: i) the onset of symptoms; ii) a positive NP RT-PCR; and iii) PDI. Cell culture experiments did not show any infectivity related to RNAemic plasmas.

Discussion

SARS-CoV-2 RNA can be detected in a small fraction of blood donors with PDI, reporting very low levels of RNA. The corresponding plasma is probably not infectious. These findings highlight the value of haemovigilance and PDI to guide blood safety strategies.

INTRODUCTION

The third decade of the 21^st century began with the emergence of coronavirus-associated disease 2019: COVID-19^1,2. Its infectious agent, SARS-CoV-2, is the third highly pathogenic coronavirus to be introduced into humans from animal reservoirs, with or without known intermediates^3,4. Within the first months of 2020, this new virus had spread worldwide, and at the moment of writing is responsible for an unprecedented pandemic. Like other respiratory-syndrome-associated coronaviruses, SARS-CoV-2 is mostly transmitted through air-borne droplets and aerosols generated by the lower and upper respiratory tracks. It is thought that transmission is also possible through the faeces^5–7 of symptomatic, and even asymptomatic, individuals, who represent an important source of contamination^8–10.

In addition to the classical routes of respiratory infections, viraemia (or more probably RNAemia, as we do not know whether RNA is associated to genuine infectious viral particles) had been reported early in the SARS-CoV-2 pandemic, on average in 10% of patients (95% CI: 5–18, random effects model)¹¹. This presence of circulating nucleic acid was generally associated with severe symptoms^11–17, and early reports stated that no RNA had been found in the plasma of asymptomatic individuals^14,18,19. Nevertheless, in April 2020, a Chinese study described two individuals with extremely low plasma viral load out of more than 7,400 blood donors who remained asymptomatic²⁰. The authors also found two positive individuals among 33 blood donors reporting fever after donation (post-donation information [PDI]). No blood products processed from RNA-positive blood had been transfused. In the aftermath of this report, the question of transmission of SARS-CoV-2 through transfusion gained new importance. A few reviews explored the rationale of such an issue, namely by comparing SARS-CoV-2 to its close relative SARS-CoV, and to a lesser extent MERS-CoV^21–24. All authors concluded that the transfusion-transmission of SARS-CoV-2 remains possible, although hypothetical.

Early on in the SARS-CoV-2 pandemic, the National Blood Service in France (EFS) implemented several measures to ensure both donor and recipient safety regarding COVID-19, in accordance with the first European recommendations. Given that the risk of SARS-CoV-2 transmission via blood had not been ascertained, the first measures were precautionary and focused on preventing an infectious donor from donating, as well as on post-donation information. The main measures to mitigate the risk of transmission by transfusion included the deferral of donors with active confirmed COVID-19 and donors exposed to SARS-CoV-2 through close contact with a confirmed case of COVID-19. Donors who have recovered from confirmed COVID-19 may donate as long as they wait at least 28 days after resolution of symptoms or on providing laboratory evidence of viral RNA clearance from the upper respiratory tract. Furthermore, blood donors were asked to report any adverse effect that could be related to a SARS-CoV-2 infection to the EFS within 15 days after donating.

In a first report²⁵, we relied on the French haemovigilance network, which collects information on adverse events occurring at any stage of the transfusion chain, to investigate 268 cases of blood donors reporting COVID-19 symptoms shortly after donation, from March to August 2020. We found 3 plasma samples positive for SARS-CoV-2 RNA showing very low viral loads that were not infectious in cell culture, including after ultracentrifugation. Here we provide these preliminary results and report data from almost one year of surveillance of SARS-CoV-2 in blood donors in France who reported SARS-CoV-2/COVID-19 related post-donation information, to confirm our first results in a large amount of samples, and compare them with demographical and clinical data.

MATERIALS AND METHODS

Study population

From February 2020 to January 2021, all blood donors who reported PDI related to a possible SARS-CoV-2 infection within 15 days after donation were included in the study. Donors were asked to report PDI as soon as possible after donation, and within 48 hours (h) if possible; however, any PDI reports made after this period were also considered. Non-transfused blood components were segregated and kept under quarantine conditions. Donors reporting PDI suspected to be related to COVID-19 were called back and completed a questionnaire to collect data including the date and type of donation (whole blood, plasma or platelets apheresis), the date and reason for the PDI (contact with a COVID-19 patient, nature of symptoms, if any, positive RT-PCR or antigenic test, serology, or chest scan), the date of diagnosis, and the evolution of symptoms and treatment, if available.

PDI donors were classified into three categories: 1) donors diagnosed with COVID-19 by RT-PCR or antigenic testing in a nasopharyngeal (NP) specimen or with clinical evidence of COVID-19 (COVID-PDI); 2) donors with symptoms that might be related to COVID-19 (possible-PDI); 3) donors reporting being in close contact to a COVID-19 patient before donation (contact case, contact-PDI). On reporting SARS-CoV-2-related PDI, one of the 2 frozen plasma samples (stored in the donation repository biobank for three years for each blood donation) was sent to the National Reference Centre for Transfusion Infectious Risks (Centre National de Référence Risques Infectieux Transfusionnels [CNR-RIT]), Paris, France, for molecular detection by RT-PCR and sequencing. When available, the plasma bag was also identified and sent to the CNR-RIT. This investigation was conducted in accordance with the Declaration of Helsinki.

Study of event chronology

We compared RNAemic and non-RNAemic individual populations regarding several duration parameters: the lapsed time between the donation and: i) the onset of symptoms (Δ1); ii) a positive NP RT-PCR (Δ2); and iii) the PDI (Δ3), as well as the lapsed time between the positive NP RT-PCR or the onset of symptoms and PDI (D4). Regarding Δ1, as 95% of people declare symptoms within 14 days²⁶ after contamination²⁷, and RNAemia has been reported in a healthy donor 40 days after the resolution of a respiratory illness²⁷, we limited the range of onset of symptoms between 40 days before and 20 days after donation. Regarding Δ2, we limited the range of positive RT-PCR between 10 days before and 20 days after donation, according to published data^28,29. Lastly, regarding Δ3, the time to PDI was limited to 28 days, according to the French deferral period for SARS-CoV-2 infection or suspicion of infection.

Molecular detection and sequencing of SARS-CoV-2

The screening and confirmation algorithm of SARS-CoV-2 RNA is described in Figure 1. Detection of SARS-CoV-2 RNA was performed using SARS-COV-2 R-GENE^® assay (Biomérieux, Craponne, France). Positive samples were confirmed using the SARS-CoV-2 real-time-PCR set up by the National Reference Centre for respiratory viruses (NRC-rt-PCR; Pasteur Institute, Paris, France) and the Procleix^® SARS-CoV-2 Assay on Panther^® System (Grifols, Barcelona, Spain) whenever the plasma bag was available (see Online Supplementary Content, Supplementary methods). Positive and inconclusive samples were subjected to Sanger sequencing using two RT-PCRs, using primers designed by the ARTIC network, targeting the spike and nucleocapsid gene, and additional primers to detect variants in repeat-reactive samples. Positive and inconclusive samples for which the plasma bag was available were also subjected to next generation sequencing using COVIDSeq Test (Illumina, San Diego, CA, USA) (see Online Supplementary Content, Supplementary methods). Samples were classified as “confirmed” when a sequence, at least partial, was obtained, “repeat-reactive” when another real-time molecular test was positive, and “suspected” if the screening assay was positive without any further positive assay (Figure 1).

Figure 1.

Screening and confirmation algorithm

Name of assays are indicated, as well as number of replicates, if appropriate. Positive samples with the screening test (R-GENE PCR1) were submitted to 2 secondary real-time (RT)-PCR assays, leading to repeat-reactive and inconclusive samples. All samples were subjected to Sanger sequencing using 2 RT-PCRs, using primers designed by the ARTIC network, targeting the spike and nucleocapsid gene, and additional primers to detect variants in repeat-reactive samples. Positive and inconclusive samples for which the plasma bag was available were also subjected to next generation sequencing using COVIDSeq Test (Illumina, San Diego, CA, USA). Samples were classified as “confirmed” when a sequence, at least partial, was obtained,” repeat-reactive” when another RT molecular test was positive, and “suspected” if the screening assay was positive without any further positive assay.

Sequence analysis

Sequences generated by direct sequencing were blasted using the Blastn online tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi) to verify that they were related to SARS-CoV-2. Deep sequencing analysis was performed using local Dragen CovidSeq Test v1.2.2 (Illumina) for demultiplexing and consensus building. Lineages were determined with Pangolin algorithm (GitHub, San Francisco, CA, USA), and clades using NextClade (GitHub). Sequences were submitted to GenBank and accession numbers MZ753785 to MZ753793 were obtained for Sanger sequencing, and SRX11368809 to SRX1136880912 for COVIDSeq (Sequence Read Archive, National Library for Biotechnology Information).

Viral culture system

Cells were propagated at 37°C (5% CO₂) in Minimal Essential Medium (MEM, Life Technologies, Carlsbad, CA, USA) supplemented with 5% foetal bovine serum (FBS), 1% penicillin/streptomycin (Life Technologies), 1% non-essential amino acids (Life Technologies), and 1 U/mL heparin. Virus growth assays were carried out in duplicate in MEM supplemented with 2% FBS. 400 μL of plasma was added to 5×10⁵ Vero E6 (ATCC, Manassas, VI, USA) cells in 12-well plates. After 1-h incubation at 37°C (5% CO₂), cells were washed with PBS before adding MEM supplemented with 2% FBS and 1 U/mL heparin. Wells were observed for cytopathic effects (CPE) by microscopy on days 3 and 7. After 7 days of incubation, a second round of culture was performed on Vero E6, using the whole volume of the supernatant of the first passage and 600 μL of MEM (2% FBS). After another 7 days under the same conditions, samples were checked for CPE, and 100 μL of culture supernatant was sampled for viral RNA extraction using the QIAamp Cador Pathogen minikit on a Qiacube HT (Qiagen, Hilden, Germany). SARS-CoV-2 detection was performed using the NRC-RT-PCR.

Statistical analysis

Data were analysed using GraphPad Prism 8.0.2 package. Population characteristics were reported as absolute numbers and percentages or means with standard deviation, as appropriate. Differences between groups regarding demographic data were tested using a χ² test for distribution, with Yates’ correction for 2-class comparison, 2-sample t-test or Welch-corrected t-test for means whenever group variances were equal or not, respectively, and ANOVA to compare means in more than two groups. p<0.05 was considered statistically significant.

Data availability

A file compiling data from the 1,092 PDI donations used in this article is available in the Mendeley Data public repository (https://data.mendeley.com//datasets/w4tyn7pfvg/1).

RESULTS

Demographics and symptoms in the study population

Between 18 February 2020 (week 8) and 18 January 2021 (week 3), the National Blood Service collected 2.58 million blood donations from 1.50 million blood donors. During this period, 2,971 PDI were related to SARS-CoV-2 or COVID-19. The reporting of COVID related-PDI over time showed peaks at week 13 and 43; these were not related to the weekly number of donations (Figure 2).

Figure 2.

Evolution of the weekly number of donations, the number of COVID-19-related post-donation information (PDI) donations, and the number of samples analysed and testing positive during the study period

A total 1,092 PDI donations (37% of the SARS-CoV-2-related PDI donations) were included in the study: 739 (68%) were COVID-PDIs, 275 (25%) were possible-PDIs, 78 (7%) were contact-PDIs. A significant difference was observed between regions (considering only Metropolitan France) regarding the proportion of reported COVID-19-related PDIs relative to the number of blood donors per regions, the Paris region reporting twice more cases than the other regions taken separately (Figure 3a). SARS-CoV-2 PDI donors were mostly women (sex ratio=0.69), between 18 and 70 years old (mean 38.1 years [95% CI: 37.3–38.9]) and no significant age difference was observed between confirmed COVID-19 (37.9 years [95% CI: 36.9–38.8]), suspected COVID-19 (38.4 years [95% CI: 36.8–39.9]) and contact cases (39.7 years [95% CI: 36.7–42.8]) (Figure 3b). In comparison, over the same period, the mean age was significantly higher in the total blood donor population (42.5 years [95% CI: 42.4–42.5]; p<0.001). There was a significant difference in age distribution between PDI blood donors and the global donor population (p<0.01), as shown by the ratio of PDI donors and all donors in five age groups (Figure 3c).

Figure 3.

Population demographic and geographic data

a. Age of blood donors (BD) reporting confirmed COVID-19 (COVID+), possible-COVID, or contact-case post-donation information (PDI). b. Distribution of age in the COVID-19-related PDI blood donor and all blood donor populations expressed as the ratio PDI BD/total BD. c. Geographic distribution of PDI blood donors (ratio to the total BD population per region) and RNAemic blood donors during the study period (ratio to the PDI donors per region).

Among 420 donors reporting information regarding symptoms, a total of 303 (72%) had symptoms: 20 (5%) from 75 days to one day before donation (median: −7.5 days, 95% CI: [−26.0/−2.0]), and should not have donated, 246 (58%) after donation (30 donors the same day as their donations, and 216 after donation (median: −2.0 days, 95% CI: [2.0/2.0]); no information regarding the onset of symptoms was available for the remaining 37 (9%). A total of 117 donors (28%) reported no symptoms at the time of call-back. Donors reporting symptoms also reported a positive PCR in 53% of cases (n=160), a negative PCR in 3% of cases (n=9, with no information regarding the date of the PCR), and no NP RT-PCR (or no information regarding a NP RT-PCR) in 44% of cases (n=134) (Table I). Symptom distribution was as follows: fever (54%), loss of smell/taste (47%), cephalalgia (40%), cough (40%), dyspnoea (19%), and others (65%), including diarrhoea, rhinorrhoea, fatigue, and joint and/or muscle pain.

Table I.

Results of NP-RT-PCR testing and the presence of symptoms at the time of testing in the 3 categories of PDI, according to the presence of SARS-CoV-2 RNAemia

Testing and symptoms	COVID-PDI (n=739)		Possible-PDI (n=275)		Contact-PDI (n=78)		Total PDI (n=1092)		Total
	RNA* pos	RNA neg	RNA pos	RNA neg	RNA pos	RNA neg	RNA pos	RNA neg
NP testing** + / symptoms +	29	130	0	1	0	0	29	131	160
NP testing − / symptoms +	0	0	0	8	0	1	0	9	9
NP testing + / no symptom	5	35	0	0	0	1	5	36	41
NP testing − / no symptom	0	0	0	0	0	3	0	3	3
NP testing ? / symptoms +	0	0	3	127	0	4	3	131	134
NP testing ? / no symptom	0	0	0	4	0	69	0	73	73
NP testing ? / symptoms ?	0	0	0	130	0	0	0	130	130
NP testing + / symptoms ?	0	540	0	1	0	0	0	541	541
NP testing − / symptoms ?	0	0	0	1	0	0	0	1	1
NP Testing +	34	705	0	2	0	1	34	708	742
NP Testing −	0	0	0	9	0	4	0	13	13
No information on testing	0	0	3	261	0	73	3	334	337
Presence of symptoms +	29	130	3	136	0	5	32	271	303
Absence of symptoms	5	35	0	4	0	73	5	112	117
No information on symptoms	0	540	0	132	0	0	0	672	672
Total	34	705	3	272	0	78	37	1,055	1,092

^*SARS-CoV-2 RNA in plasma (RNAemia);

^**NP testing: Nasopharyngeal RT-PCR or antigenic testing;

−: negative NP testing; +: positive (NP testing)/presence (symptoms); ?: no information.

According to reported information, among the 1,092 investigated PDI-blood donors, 742 (68%) were diagnosed by NP RT-PCR screening (n=723) or antigenic testing (n=19) between 175 days before and 61 days after donation, and 13 (1%) donors tested negative by NP RT-PCR (9 with symptoms and 4 contact cases) (Table I). For 337 (31%) donations, no information regarding NP testing was available. In 258 donations for which these data were available, the median interval between the positive NP RT-PCR or the onset of symptoms (whichever came first) and PDI (Δ4) was 2 days, and 75% of the PDI were reported within 5 days.

Investigation of the post-donation information samples testing positive for SARS-CoV-2 RNA

Thirty-seven donations (3.4%) were found initially reactive with R-GENE PCR1, with Ct values between 36.5 and 42.0 (median=38.9, 95% CI: [38.6;39.1]), suggesting very low viral loads (Table II). Of these 37, 34 (92%) were COVID-PDIs and 3 (8%) were possible-PDIs. No contact-case was found positive for SARS-CoV-2 RNA. Thirty-two (86%) of the 37 reported symptoms while the other 5 (14%) remained asymptomatic. Most RNAemic samples came from the south (41%; n=15), followed by the Paris region (24%; n=9); however, northern regions proportionally reported more cases than the Paris region (Figure 3a). Eleven of 37 samples (30%) tested positive with another RT-PCR assay (Ct between 36.4 and 40.5, mean=38.1, 95% CI: [37.1;39.0]; PCR2, n=1; NRC-RT-PCR, n=9). Of the 26 donations for which the plasma bag was available, 9 (35%) also tested positive with the Procleix^® SARS-CoV-2 assay (TMA [Grifols]). Only 7 of the 37 samples (19%) were confirmed by Sanger sequencing, with one or more pairs of primers: 6 samples tested positive with ARTIC75, 4 with ARTIC76 and ARTIC77, and 2 with ARTIC95 (Table II). Deep sequencing (COVIDSeq [Illumina]) only allowed confirmation of 4 samples that were also confirmed by SANGER sequencing. No mutations related to variants were found. Lastly, on the 26 plasma units available for virus isolation, none of the cell culture experiments yielded a virus growth after 2 rounds (14 days).

Table II.

Results of molecular a^* virus culture investigations in donations screened positive for SARS–CoV–2 RNA

PDI number	R–GENE PCRs		NRC PCR		Procleix SARS–CoV–2	SANGER				NGS	Confirmed		Virus isolation
	PCR1 (N; Ct)	PCR2 (E; Ct)	IP2 (Ct)	IP4 (Ct)		ARTIC 75 (S)	ARTIC 76 (S)	ARTIC 77 (S)	ARTIC 95 (N)	COVID–Seq	by another real–time assay	by sequencing
1	37.9	−	37.0/37.1	−	2/5	+	−	−	+	−	yes	yes	−
2	38.0	38.4	−	−	−	−	+	−	−	−	yes	yes	−
3	37.4	−	37.4	−	−	+	−	−	−	+	yes	yes	−
4	36.9	−	36.9/38.4	36.7/39.5	2/3	+	+	+	−	+	yes	yes	−
5	36.5	−	38.1	39.7	3/5	+	−	+	+	+	yes	yes	−
6	38.9	−	38.2/38.5	39.0	*	*	*	−	−	−	yes	*	*
7	38.8	−	−	−	−	*	*	−	−	−	*	*	−
8	37.9	−	−	−	−	*	*	−	−	−	*	*	−
9	41.3	−	−	−	−	*	*	−	−	−	*	*	−
10	37.8	−	38.4	40.5	1/3	−	−	−	−	−	yes	*	−
11	39.7	−	−	−	−	*	*	−	−	−	*	*	−
12	38.8	−	−	−	−	*	*	−	−	−	*	*	−
13	39.1	−	−	−	*	*	*	−	−	−	*	*	*
14	37.4	−	−	−	*	*	*	−	−	−	*	*	*
15	38.2	−	−	−	*	*	*	−	−	−	*	*	*
16	39.5	−	−	−	*	*	*	−	−	−	*	*	*
17	39.1	−	−	−	*	*	*	−	−	−	*	*	*
18	39.8	−	−	−	*	*	*	−	−	−	*	*	*
19	39.0	−	−	−	−	*	*	−	−	−	*	*	−
20	39.9	*	−	39.3	2/5	+	+	+	−	−	yes	yes	−
21	41.4	−	−	−	−	*	*	−	−	−	*	*	−
22	38.6	−	−	−	−	*	*	−	−	−	*	*	−
23	39.0	−	−	−	−	*	*	−	−	−	*	*	−
24	39.1	*	−	39.0	*	*	*	−	−	−	yes	*	*
25	39.4	−	−	−	−	*	*	−	−	−	*	*	−
26	39.1	*	−	37.3/36.8/38.0	3/3	+	+	+	−	+	yes	yes	−
27	39.5	−	−	−	−	*	*	−	−	−	*	*	−
28	39.4	−	−	−	−	*	*	−	−	−	*	*	−
29	40.2	−	−	−	*	*	*	−	−	−	*	*	*
30	37.7	−	−	−	*	*	*	−	−	−	*	*	*
31	38.6	−	−	−	−	*	*	−	−	−	*	*	−
32	42.0	−	−	−	−	*	*	−	−	−	*	*	−
33	37.2	−	−	−	−	*	*	−	−	−	*	*	−
34	40.1	−	−	−	*	*	*	−	−	−	*	*	*
35	38.1	−	−	−	−	*	*	−	−	−	*	*	−
36	38.8	−	−	−	1/3	*	*	−	−	−	yes	*	−
37	38.7	−	−	−	−	*	*	−	−	−	*	*	−

−: negative; +: positive; N: nucleocapsid gene; E: small envelope protein gene; S: spike gene;

^*not done (nothing specified).

Of the 303 blood donors reporting symptoms compatible with COVID-19, 271 were non-RNAemic and 32 RNAemic. Among them, 233 non-RNAemic and 31 RNAemic blood donors reported a date of symptoms onset in a range compatible with the PDI. There was no statistical difference in the median lapsed time between donation and symptoms (Δ1) between the non-RNAemic (median Δ1=2.0 days, 95% CI: [2.0;2.0]) and the RNAemic group (Δ1=2.0 days, [95% CI: 1.0–2.0 days]) (Figure 4a). Likewise, 397 non-RNAemic and 30 RNAemic blood donors reported a positive RT-PCR or antigenic test in an upper respiratory sample, in a range compatible with the PDI (10 days before and 20 days after). There was no statistical difference in the median lapsed time between donation and a positive RT-PCR in upper respiratory samples (Δ2) between the non-RNAemic (median Δ2=4.0 days, 95% CI: [3.0–4.0]) and the RNAemic (Δ2=3.5 days, 95% CI: [3.0–5.0]) groups (Figure 4b). For the 350 donations with a known date of PDI, the median lapsed time between donation and PDI (Δ3) was 4 days, with 75% of PDI declared within 7 days; there was no difference in the median Δ3 between the 315 non-RNAemic and the 35 RNAemic blood donors for which this information was available (median Δ3=4.0 days, 95% CI: [3.0–5.0] for both groups) (Figure 4c). No significant difference was found between RNAemic and non-RNAemic donors regarding the distribution of symptomatic (n=32 and 271, respectively) and asymptomatic (n=5 and 112, respectively) individuals (p=0.06, Yates’ correction) (Table I). Lastly, no significant difference was observed between RNAemic and non-RNAemic individuals regarding the lapsed time between a positive NP RT-PCR or the onset of symptoms and PDI (median Δ4=2.0 days, 95% CI: [1.0–2.0] and [2.0–3.0], respectively). To study the possibility that missing data differ systematically from available data, we conducted a contingency analysis regarding gender and age groups (demographic data available for all donors) for the Δ1, Δ2, Δ3, and Δ4 time parameters. No significant difference was observed for any of these parameters and the aforementioned hypothesis was, therefore, excluded (see Online Supplementary Content, Table SII).

Figure 4.

Lapsed time between donation and (a) symptoms (Δ1); (b) a positive test in the upper respiratory samples (Δ2); and (c) post-donation information (PDI) (Δ3), and (d) between a positive nasopharyngeal (NP) real-time (RT)-PCR or the onset of symtoms and PDI (Δ4)

Median and 95% Confidence Interval (95% CI) of median are indicated by the black and gray horizontal bars, respectively.

Overall, when considering the whole PDI population, the positivity rate is 3.4% (37 of 1,092). If we consider that PDI are valid if reported within 35 days after donation (7 days of symptoms and 28 days of deferral period), 352 donations would be considered as a denominator, among which 35 were found to be SARS-CoV-2 positive in plasma. This leads to a worst-case scenario positivity rate of 10%.

Lookback investigations

From the positive donations, two red blood cell units (RBU) were transfused: one to a COVID-19 positive patient²⁵, the second to a patient who remained asymptomatic within two months after transfusion (no RT-PCR testing). Sixteen inactivated pooled platelet concentrates (PPC) were transfused to 16 patients, and one PPC to 2 patients; none of the recipients showed COVID-19-related symptoms after transfusion.

DISCUSSION

The PDI data collected from 1,092 donations during the first year of the SARS-CoV-2 epidemic in France showed that the donors involved were significantly younger than the general blood donor population and more often lived in the Paris region. A possible explanation for this is that young donors felt less exposed, or at least less at risk of exposure, and were probably less reluctant to travel to a blood collection centre compared to older people. In addition, the Paris region was one of the most affected by COVID-19. The number of SARS-COV-2-related PDI developed in line with the epidemic waves in France, with peaks at weeks 13 and 43. Of the 1,092 donations studied, between 3 (37 of 1,092; best-case scenario) and 10% (35 of 350; worst-case scenario) tested positive for viral RNA, with 92% of corresponding donors (34 of 37) reporting a confirmed COVID-19 during the post-donation interview. This is in accordance with data from the USA (M. Busch, reported at the ISBT Conference, 2021). Excluding individuals with serious COVID-19, for which RNAemia is correlated to the severity of the disease, RNAemia in asymptomatic individuals could be the manifestation of a fleeting systemic infection in the early stages of the viral invasion.

There was no duration parameter (lapsed time between donation and other events) that could distinguish RNAemic and non-RNAemic donors. The highest number of positive samples was found in the south and the north, despite the fact that the Paris region reported more PDI than the other regions. Nevertheless, this trend must be interpreted with caution due to the small number of reactive samples. In our dataset, the majority of PDI (75%) related to SARS-CoV-2 infection was given within 5 days after a positive NP RT-PCR or the onset of symptoms, underlining the awareness of blood donors of the need to provide post-donation information. Most RNAemic donors had symptoms, and only 5 remained asymptomatic at the time of call-back, in the same range as non-RNAemic donors, even though data are not complete for the latter. Notably, symptoms found in blood donors were those reported in non-severe patients, as described in two meta-analysis studies (46,959 and 3,420 patients)^30,31 and one series with 352 COVID-19+ patients³². Dyspnoea was less often described than in clinical series (20%), probably because this is more likely to be found in individuals with a more severe outcome.

As already reported in our first study²⁵, positive samples showed high Cts (>36.5) in only one target of duplex RT-PCRs with a non-repeat reactive behaviour, confirming very low viral loads in individuals who were asymptomatic at the time of screening. Only 27% of samples could be confirmed by another RT-PCR/TMA assay, and 19% by sequencing. Deep sequencing with multiplex targeted amplification did not show better sensitivity when compared with Sanger sequencing. This non-repeat behaviour could be explained by extremely low viral loads, by fragmented viral RNA in the blood stream, or by non-specific reactivity. No variants of concern were found in the analysed sequences, reflecting the fact that positive samples correspond to the early period of the pandemic. Lastly, as already shown in blood donors²⁵ and in patients¹¹, no infectivity was associated to RNAemia in laboratory in vitro conditions. It should be noted that pathogen inactivation is used on all platelet concentrates in France and is reported to be active against SARS-CoV^33,34, therefore platelet concentrates that were transfused early after donation before PDI can be considered non-infectious.

Almost two years after the first cases of COVID-19 were discovered, the literature is still scarce regarding SARS-CoV-2 and blood safety, even though the transfusion community generally believes that this novel agent represents no real threat via transfusion. Two reviews remind us that four major factors are needed for a pathogen to cause a transfusion-transmitted infection (TTI): 1) the presence of the agent in an infectious form in the blood of asymptomatic individuals; 2) the survival of the agent during the processing of blood (including the pathogen-reduction technologies) and the storage of the blood components; 3) the transmission of the agent via the hematogenous route; and 4) a clinically apparent disease induced in the recipient of contaminated blood products^21,22. If the Chinese study²⁰ validated the presence of SARS-CoV-2 in donor plasma, the three other factors remain hypothetical, except as far as the pathogen inactivation is concerned^33,34. To date, 13 studies have reported prospective or retrospective screening of SARS-CoV- 2 RNA in blood donations (n=6)^{19,20,27,35–37}, and/or the investigations of PDI related to COVID-19 and/or a positive SARS-CoV-2 RT-PCR in respiratory samples (n=8)^{18,20,25,38–42} (see Online Supplementary Content, Table SII). In total, 25 blood components (16 platelet concentrates and 9 RBU) were produced with blood from pre- or asymptomatic individuals and transfused to 22 recipients. In 9 studies, plasma was tested for SARS-CoV-2 RNA, and 6 reported SARS-CoV-2 RNAemia, corresponding to 11 positive individual plasma samples and 3 positive minipools (no individual testing). None of the 22 recipients developed symptoms after transfusion, and 16 recipients also tested negative for SARS-CoV-2 RNA in NP swabs (Online Supplementary Content, Table SII). Only one recipient tested positive two days after transfusion, but neither the donor nor the corresponding plasma was tested for SARS-CoV-2 at the time of the PDI. Finally, one study also reported 2 cases of allogeneic hematopoietic stem cell transplantation, with cells harvested from donors who tested positive for SARS-CoV-2 by RT-PCR on NP swabs at the time of cell collection. None of the recipients developed COVID-19, and all tests remained negative over a 4-week follow-up period⁴³. In addition, few data are available on the real infectious power of SARS-CoV-2 in blood. One meta-analysis summarised results on RNAemia in 1,512 patients from 22 relevant studies¹¹. Levels of viral loads are heterogeneous and generally low, and are associated with increased risk of critical disease and death. The authors also performed blood infectivity testing using PCR-positive sera, and did not observe any cytopathic effect, suggesting the absence of relatedness between the results of molecular testing in plasma and infectivity. To date, there has been no large-scale study reporting infectivity of SARS-CoV-2-RNA-positive plasma in blood donors.

The French haemovigilance network allowed us to gather epidemiological and virological data on more than 1,000 donations associated with PDI. One limitation of our study is that only one-third (1,092 of 2,971) of SARS-CoV-2-related PDI could be investigated. This was due to several reasons: 1) the need to urgently implement new PDI measures during the first epidemic wave; 2) the interruption of the biological investigation of contact cases due to the low probability to find a positive in this group of donors (28% of the total SARS-CoV-2-related PDI; data not shown); 3) the limited testing of individuals by NP-RT-PCR at the beginning of the pandemic, as shown in Figure 2 by the artifactual shift from a higher number of possible-PDI (not biologically confirmed) in the first wave to a majority of PDI with a confirmed infection (COVID-PDI) over the second wave; 4) the delay in notification of some PDI, as donors sometimes disclosed COVID-19-related symptoms at the following donation, and these donations were not investigated; and 5) the rate of donors who were called back and the accuracy of their answers to the questionnaire that impaired the comprehensiveness of the dataset.

CONCLUSIONS

After one year since the beginning of the pandemic, no cases of SARS-CoV-2 transmission by transfusion have been reported; this supports the in vitro data and previous data on SARS- and MERS-CoV^21–24. The data generated here using blood donation subjected to PDI corroborated the findings in patients and tend to offer reassurance as to the safety of blood with regards to SARS-CoV-2, based on pre-donation selection, without the need for further nucleic acid testing. In the face of the huge increase in SARS-CoV-2 infections, in agreement with the international recommendations, COVID-related PDI was rapidly implemented in France as a precautionary measure to mitigate the hypothetical risk of the transmission of SARS-CoV-2 by transfusion. The difficulty in ensuring that donor-reported information is truly related to COVID-19, and the incentive to report any post-donation event, probably led to the number of COVID-PDI being over-estimated and to many blood components being mistakenly discarded. Nevertheless, thanks to the haemovigilance network established in France over many years, we have been able to provide new insights into the infection, particularly with regard to the presence of viral RNA in plasma.