Content uploaded by Jessica Rose
Author content
All content in this area was uploaded by Jessica Rose on Oct 24, 2023
Content may be subject to copyright.
Content uploaded by David M Wiseman
Author content
All content in this area was uploaded by David M Wiseman on Oct 20, 2023
Content may be subject to copyright.
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 1
DNA fragments detected in monovalent and bivalent
1
Pfizer/BioNTech and Moderna modRNA COVID-19 vaccines
2
from Ontario, Canada: Exploratory dose response
3
relationship with serious adverse events.
4
5
David J. Speicher1, Jessica Rose2, L. Maria Gutschi3, David Wiseman4,
6
Kevin McKernan5
7
8
1Department of Pathobiology, University of Guelph, Guelph, Ontario, Canada
9
2Independent Researcher, Ontario, Canada ORCID 0000-0002-9091-4425
10
3Pharmacy Consultant, Ottawa, Ontario, Canada
11
4Synechion, Inc., Dallas, Texas, USA ORCID 0000-0002-8367-6158
12
5Medicinal Genomics, Beverly, MA, USA ORCID 0000-0002-3908-1122
13
14
Corresponding Author: Dr. David J. Speicher
15
University of Guelph
16
50 Stone Rd E, Guelph, ON, N1G 2W1
17
speicher@uoguelph.ca
18
ORCID 0000-0002-1745-3263
19
20
Keywords: COVID-19, vaccines, DNA contamination, impurity, residual DNA, modRNA,
21
mRNA, adverse events
22
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 2
Abstract
23
Background: In vitro transcription (IVT) reactions used to generate nucleoside modified
24
RNA (modRNA) for SARS-CoV-2 vaccines currently rely on an RNA polymerase
25
transcribing from a DNA template. Production of modRNA used in the original Pfizer
26
randomized clinical trial (RCT) utilized a PCR-generated DNA template (Process 1). To
27
generate billions of vaccine doses, this DNA was cloned into a bacterial plasmid vector
28
for amplification in Escherichia coli before linearization (Process 2), expanding the size
29
and complexity of potential residual DNA and introducing sequences not present in the
30
Process 1 template. It appears that Moderna used a similar plasmid-based process for
31
both clinical trial and post-trial use vaccines. Recently, DNA sequencing studies have
32
revealed this plasmid DNA at significant levels in both Pfizer-BioNTech and Moderna
33
modRNA vaccines. These studies surveyed a limited number of lots and questions remain
34
regarding the variance in residual DNA observed internationally.
35
Methods: Using previously published primer and probe sequences, quantitative
36
polymerase chain reaction (qPCR) and Qubit® fluorometry was performed on an
37
additional 27 mRNA vials obtained in Canada and drawn from 12 unique lots (5 lots of
38
Moderna child/adult monovalent, 1 lot of Moderna adult bivalent BA.4/5, 1 lot of Moderna
39
child/adult bivalent BA.1, 1 lot of Moderna XBB.1.5 monovalent, 3 lots of Pfizer adult
40
monovalent, and 1 lot of Pfizer adult bivalent BA.4/5). The Vaccine Adverse Events
41
Reporting System (VAERS) database was queried for the number and categorization of
42
adverse events (AEs) reported for each of the lots tested. The content of one previously
43
studied vial of Pfizer COVID-19 vaccine was examined by Oxford Nanopore sequencing
44
to determine the size distribution of DNA fragments. This sample was also used to
45
determine if the residual DNA is packaged in the lipid nanoparticles (LNPs) and thus
46
resistant to DNaseI or if the DNA resides outside of the LNP and is DNaseI labile.
47
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 3
Results: Quantification cycle (Cq) values (1:10 dilution) for the plasmid origin of
48
replication (ori) and spike sequences ranged from 18.44 - 24.87 and 18.03 - 23.83 and
49
for Pfizer, and 22.52 – 24.53 and 25.24 – 30.10 for Moderna, respectively. These values
50
correspond to 0.28 – 4.27 ng/dose and 0.22 - 2.43 ng/dose (Pfizer), and 0.01 -0.34
51
ng/dose and 0.25 – 0.78 ng/dose (Moderna), for ori and spike respectively measured by
52
qPCR, and 1,896 – 3,720 ng/dose and 3,270 – 5,100 ng/dose measured by Qubit®
53
fluorometry for Pfizer and Moderna, respectfully. The SV40 promoter-enhancer-ori was
54
only detected in Pfizer vials with Cq scores ranging from 16.64 – 22.59. In an exploratory
55
analysis, we found preliminary evidence of a dose response relationship of the amount of
56
DNA per dose and the frequency of serious adverse events (SAEs). This relationship was
57
different for the Pfizer and Moderna products. Size distribution analysis found mean and
58
maximum DNA fragment lengths of 214 base pairs (bp) and 3.5 kb, respectively. The
59
plasmid DNA is likely inside the LNPs and is protected from nucleases.
60
Conclusion: These data demonstrate the presence of billions to hundreds of billions of
61
DNA molecules per dose in these vaccines. Using fluorometry, all vaccines exceed the
62
guidelines for residual DNA set by FDA and WHO of 10 ng/dose by 188 – 509-fold.
63
However, qPCR residual DNA content in all vaccines were below these guidelines
64
emphasizing the importance of methodological clarity and consistency when interpreting
65
quantitative guidelines. The preliminary evidence of a dose-response effect of residual
66
DNA measured with qPCR and SAEs warrant confirmation and further investigation. Our
67
findings extend existing concerns about vaccine safety and call into question the
68
relevance of guidelines conceived before the introduction of efficient transfection using
69
LNPs. With several obvious limitations, we urge that our work is replicated under forensic
70
conditions and that guidelines be revised to account for highly efficient DNA transfection
71
and cumulative dosing.
72
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 4
Introduction
73
To produce large amounts of modified RNA (modRNA) vaccine for generalized use, Pfizer
74
changed its manufacturing process (Process 1) used to produce material for the
75
randomized clinical trial (RCT)1 to a process (Process 2) similar to the one already being
76
used by Moderna. The SARS-CoV-2 spike sequence was cloned into a plasmid
77
containing a bacterial origin of replication (generically termed ori) active in Escherichia
78
coli. This plasmid (7,824 base pairs (bp) and 6,777 bp for Pfizer and Moderna,
79
respectively) also contains an aminoglycoside phosphotransferase gene (Neo/Kan) that
80
allows cost effective bacterial replication in a broth containing kanamycin and a doubling
81
of plasmid copy number every 30 minutes at 37oC. The E. coli cells are then harvested
82
and lysed. DNA is extracted and linearized with the restriction enzyme Eam1104I. This
83
linear DNA then acts as the template for T7 RNA Polymerase in vitro transcription (IVT)
84
in the presence of N1-methyl-pseudouridine. After the IVT, DNA is hydrolyzed, reducing
85
its prevalence in the final drug product. Documents leaked from the European Medicines
86
Agency (EMA) and cited in the British Medical Journal2 noted that residual DNA in
87
modRNA products made by this process could vary significantly3.
88
89
McKernan et al. performed next-generation RNA sequencing of these vaccines and,
90
unexpectedly, found evidence of DNA derived from the expression plasmids used during
91
manufacturing.4 McKernan et al. then developed a quantitative polymerase chain reaction
92
(qPCR) method towards the DNA contamination with primers targeting shared sequences
93
in both Pfizer and Moderna vaccines.4 Additionally, McKernan et al., found SV40
94
promoter-enhancer-ori, and SV40 polyA signal sequences in the Pfizer vaccines. To
95
investigate the generalizability of these findings to other lots of vaccines, we obtained 24
96
unopened expired vials (8 Pfizer and 16 Moderna) and three vials of in-date remnants of
97
Moderna XBB.1.5 COVID-19 vaccines that had been distributed in Ontario, Canada and
98
examined them via Qubit® fluorometry and qPCR targeting spike, plasmid ori, and the
99
SV40 promoter-enhancer-ori. We then queried the Vaccine Adverse Event Reporting
100
System (VAERS) for any adverse events (AEs), including serious AEs (SAEs), associated
101
with these lots.5 We also extended the observations of an earlier work (McKernan et al.4)
102
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 5
by studying the size distribution of DNA fragments as well as the DNasel sensitivity of the
103
vaccine to determine whether the residual DNA is packaged in the LNPs.
104
105
For the purpose of this study, we are using the terms "residual DNA," "DNA mass," (or
106
similar) rather than "impurity" or "contamination" as a discussion of these regulatory terms
107
is beyond the scope of this paper.
108
109
Methods
110
111
COVID-19 Vaccines Tested
112
Expired unopened vials of Pfizer-BioNTech BNT162b2 (n=8) and Moderna Spikevax
113
mRNA-1273 (n=16) were obtained from various pharmacies in Ontario, Canada (Figure
114
1). Three vials of in-date remnants of the same lot of Moderna XBB.1.5 vaccine were also
115
obtained. In total, 12 lots were surveyed across 27 mRNA vials: 5 lots of Moderna
116
child/adult monovalent, 1 lot of Moderna adult bivalent BA.4/5, 1 lot of Moderna child/adult
117
bivalent Wuhan-BA.1, 1 lot of Moderna XBB.1.5 monovalent, 3 lots of Pfizer adult
118
monovalent, and 1 lot of Pfizer adult bivalent Wuhan-BA.4/5 vaccines. An unopened
119
sterile injectable vial of alprostadil 66 mcg/mL in combination with papaverine 21.7mg/mL
120
and phentolamine 1 mg/mL (TriMix) was used as the negative control. The unopened
121
vials were untampered as they had intact flip-off plastic caps with printed lot numbers and
122
expiration dates. Vials had been stored in a purpose-built vaccine unit at +2-8oC in the
123
pharmacies and were transported in insulated containers with frozen gel packs and
124
placed in the testing laboratory fridge within 5 hours. Only one Moderna vial did not have
125
a printed expiration date but had a QR code that required scanning by a pharmacist. The
126
Moderna XBB.1.5 vials were similarly stored by the pharmacy. Vials were removed from
127
the refrigerator, warmed for ~20 minutes, and administered by the pharmacist to patients
128
over ~30 minutes. The remnant vials were placed in an insulated container with frozen
129
gel packs and transported to the testing laboratory fridge within 12 hours.
130
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 6
131
Figure 1. Vials of COVID-19 vaccine from Ontario, Canada: (A) Pfizer/BioNTech
132
BNT162b2 adult monovalent and bivalent; Moderna Spikevax mRNA-1273 (B) adult
133
monovalent XBB.1.5, (C) child/adult monovalent, (D) child/adult bivalent Wuhan-BA.1
134
and (E) child/adult bivalent Wuhan-BA.1 and adult Wuhan-bivalent BA.4/5.
135
136
qPCR Analysis of Spike, ori, and the SV40 Promoter-Enhancer-ori DNA
137
Each vial was tested by quantitative PCR (qPCR) for the presence of plasmid derived
138
SARS-CoV-2 spike, ori, and the SV40 promoter-enhancer-ori DNA. Spike and plasmid
139
ori were tested in duplicate with PCR primers targeting sequences shared by the Moderna
140
and Pfizer expression plasmids (Table 1). The uniplex SV40 Enhancer assay was
141
designed to amplify the nuclear targeting sequence unique to the Pfizer vector6. In brief,
142
the qPCR assays used 1 µL from each vial directly added to 17.8 µL of master mix. qPCR
143
kits were sourced from Medicinal Genomics (Part# 420201, Beverly, USA) with the
144
master mix containing 8.8 µL reaction consisting of 3.8 µL polymerase enzyme, 0.8 µL
145
reaction buffer and 1.0 µL of Primer-Probe mix, and 12.2 µL of ddH20. The Primer-Probe
146
mix was assembled using 12.5 µL 100 µM ori probe, 12.5 µL of 100 µM spike probe, 25
147
µL of 100 µM spike forward primer, 25 µL of 100 µM spike reverse primer, 25 µL of 100
148
µM ori forward primer, 25 µL 100 µM ori reverse primer, and 75 µL of ddH20.
149
150
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 7
Spike and ori qPCR assays used a synthetic gDNA control (gBlock, Integrated DNA
151
Technologies (IDT), San Diego, USA) of known concentration to generate a 10-fold serial
152
dilution derived calibration curve. The SV40 enhancer gBlock failed initial synthesis and
153
a standard curve could not be produced.
154
155
Table 1. Primer and probe sequences targeting spike, ori, and the SV40 promoter.
156
157
158
Cycling was performed on a QuantStudio 3 (ThermoFisher Scientific, Waltham, USA) with
159
an initial denaturation of 95°C for 3 minutes followed by 35 cycles of 95°C for 10 seconds
160
and 65°C for 30 seconds. Cq conversion to ng/µL was calculated using the QuantStudio
161
software v2.7.0 (ThermoFisher Scientific). Amplicon mass, as determined with the New
162
England BioLabs DNA calculator,7 and length (105 bp for ori, 114 bp for spike) were used
163
to estimate the total nanograms (ng) of DNA present by adjusting for the length of the
164
plasmids (7,824bp for Pfizer and 6,777bp for Moderna). Copy number per dose was
165
adjusted for the volume of each intramuscular vaccine injection (300 µL for Pfizer and
166
500 µL for Moderna). Serial dilutions were performed on the three Pfizer lots that showed
167
the highest residual DNA concentration. to investigate PCR inhibition by the LNPs since
168
qPCR was performed directly without any treatment or extraction.
169
170
Qubit® fluorometry quantitation
171
AccuGreen® HS fluorometric reagents (AccuGreen #99820 and DNA Quantification
172
Buffer #99979) and standards were acquired from Biotium (San Francisco, USA) for
173
Qubit® analysis (ThermoFisher Scientific). Fluorometric reagents (190 µL of a stock made
174
from 995 µL HS Buffer and 5 µL 200X AccuGreen dye) were vortexed with 10 µL of
175
vaccine. These samples were heated to 95°C for 8 minutes and 4°C for 5 minutes to
176
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 8
disrupt the LNPs and enable Fluorometric Dyes to access the DNA. Samples were read
177
following the manufacturer’s instructions on a Qubit 3.0 Fluorometer. Qubit fluorometry
178
and qPCR data were compared.
179
180
Vaccine Adverse Event Reporting System (VAERS) Data
181
The VAERS database was analyzed using the Language and Environment for Statistical
182
Computing package in R,8 and included data spanning December 17, 2020 through
183
October 6, 2023 The VAERS data is available for download in three separate comma
184
separated values (csv) data files representing: i) general data for each report; ii) the
185
reported AEs or ‘symptoms’, and iii) vaccine data including vaccine manufacturer and lot
186
number.5 A VAERS ID number is assigned to preserve confidentiality when a report is
187
filed. To assess the AEs related to a particular vaccine, it is necessary to merge the three
188
data files using the VAERS IDs as a linking variable. For this study, since we are
189
interested in the COVID-19 products, only COVID-19 vaccine type (COVID19-1
190
(monovalent) and COVID19-2 (bivalent)) were included. Other relevant variables included
191
VAERS ID*, vaccine lot (VAX_LOT), vaccine manufacturer (VAX_MANU),
192
hospitalizations (HOSPITAL) and deaths (DIED). Data were grouped by vaccine lot and
193
the total number of AE and SAE reports were counted. SAE reports included deaths,
194
hospitalizations, emergency room visits, disability reports, birth defects and life-
195
threatening reports, and individual MedDRA coded AEs, such as total deaths per lot, were
196
also counted.
197
198
The various limitations of VAERS are widely acknowledged, for example by FDA9, and
199
include underreporting, misreporting, spontaneous reporting, and the inability to infer
200
causality. Nevertheless, to explore a possible dose-response relationship between
201
residual DNA content and SAEs, we used the ratio of the number of SAE reports to the
202
total number of AEs (“SAE reporting ratio” = SRR) as a proxy for a possible toxicological
203
effect. We used the total number of AEs reported by lot as a proxy for the total number of
204
doses administered, since this denominator is difficult to estimate. This principle is used
205
by the CDC in disproportionality signal analysis (DSA) to identify safety signals using the
206
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 9
Proportional Reporting Ratio (PRR)10 The PRR, as devised by Evans et al., is a useful
207
tool in pharmacovigilance with known limitations.11
208
209
It must be noted that although VAERS is a USA-based database, it accepts reports from
210
around the world. Certain categories of AEs that are reported to manufacturers outside
211
the USA, must be reported to the VAERS database. Differences in propensity for
212
underreporting as well as mandatory reporting imposed on manufacturers or medical
213
professionals within and outside the USA may introduce confounding to the estimation of
214
the SRR. Accordingly, for our exploratory dose-response analysis we only used VAERS
215
data originating outside the USA to reduce this confounding. Additionally, we have noted
216
some discrepancies in data obtained through the downloaded version of the VAERS
217
dataset, and those obtained using the VAERS WONDER front-end web-based interface
218
(https://wonder.cdc.gov/controller/datarequest/D8). We used the downloaded version as
219
it provides greater detail than the web version. The SRR was then plotted against levels
220
of DNA found in the vials to identify any association between residual DNA levels and the
221
frequency of reports of serious adverse events.
222
223
Where more than one vial was available in any lot, the average mass of residual DNA per
224
dose for that lot was used. Zero values of SRR for any given lot were only plotted if one
225
or more AEs had been identified worldwide, signifying that that lot had actually been
226
deployed. The curves were plotted on a logarithmic axis and a trend line drawn using the
227
linear function within Microsoft® Excel.
228
229
Oxford Nanopore Sequencing
230
In a separate experiment using previously sequenced vaccine4 (Pfizer children’s
231
monovalent Lot# FL8095), DNA fragment size distributions were estimated using an
232
Oxford Nanopore Flongle (R.10.4.1, Oxford Nanopore Technologies (ONT), New York,
233
USA) and the Oxford Nanopore Ligation sequencing kit (SQK-LSK114) according to the
234
manufacturer’s instructions. Reads were mapped to NCBI OR134577.1 with the Burrow-
235
Wheeler Aligner with maximum exact matches (BWA-MEM).12 ONT sequencing read
236
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 10
length is unlimited, but the DNA isolation procedures can bias the length of the molecules
237
captured in the ONT ligation reaction. Single molecule reads were counted and binned
238
according to their mapped read length with BWA-MEM.
239
240
Nuclease sensitivity of the vaccines
241
The same vial (Pfizer Lot# FL8095) was used to assess DNasel sensitivity of the vaccine
242
by determining if the DNA contamination is packaged in the LNP and thus resistant to
243
DNaseI or if the DNA resides outside of the LNP and is DNaseI labile.
244
245
Nuclease protected DNA was estimated by treating 20 µL of the vaccine with 2.5 µL of
246
DNaseI-XT (2 units/µL, NEB#M0570S, New England BioLabs Inc, Ipswich, USA), 2.5 µL
247
of Grim Reefer 10X buffer (Medicinal Genomics #420123-125) and incubating at 37°C for
248
30 minutes. For the control, 2.5 µL of ddH20 was used instead of the DNaseI-XT. The
249
DNaseI-XT reaction was chemically arrested using 2.5 µL of MGC lysis buffer (Medicinal
250
Genomics #420001). After the DNaseI chemical kill step, a qPCR amplifiable internal
251
control DNA was spiked-in to verify that the DNaseI-XT had been fully inactivated
252
(Medicinal Genomics #420123-125).
253
254
After spiking in the DNaseI inactivation control, 54 µL of SenSATIVAx magnetic beads
255
(Medicinal Genomics) were used to purify DNA from the DNaseI-XT assay and the
256
DNaseI-XT negative control samples. The magnetic beads were pipette mixed 10 times
257
with the sample, incubated at room temperature for 5 minutes, magnetically separated
258
and washed twice with 70% v/v ethanol. The ethanol was removed, and the beads dried
259
for 2 minutes at room temperature. Samples were eluted in 30 µL of ddH20 and 1 µL of
260
eluate was examined by qPCR for spike and ori in an 18.8 µL reaction. An additional
261
DNaseI inactivation control primer and probe (0.5 µL in CY5) were added to the assay for
262
a total of 19.3 µL reaction.
263
264
Results
265
An 8-log serial dilution standard curve was used to calibrate sample Cq values and
266
generated R2 values of 0.998 and 0.999 for spike and ori amplicons, respectively. PCR
267
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 11
efficiency was 99.8% and 94.7% for spike and ori, respectively (Figure 2). On all plates,
268
negative controls and no template (ddH20) controls (NTC) were tested in triplicate and
269
found to be negative.
270
271
272
Figure 2. Calibration curves of Spike (red) and ori (blue) diluted 10-fold and tested by
273
qPCR.
274
For individual vials, qPCR on Pfizer amplified at a similar time for spike, ori, and SV40
275
enhancer-promoter-ori (ΔCq 1.48 ± 0.32) (Figure 3). Apart from Pfizer lot: FX4343, the
276
inter vial difference was small for both Pfizer (spike Cq 16.91 ± 0.52; ori Cq 16.91 ± 1.07;
277
SV40 promoter-enhancer-ori Cq 15.46 ± 2.02) and Moderna (spike Cq 20.35 ± 0.65; ori
278
Cq 25.34 ± 1.47) (values were based on the undiluted vials contents) (Table 2, Figure 4)
279
280
However, for all Moderna vials, except lot AS0467D, ori consistently amplified Cq 5-6
281
later than spike. The SV40 promoter-enhancer-ori was detected in all Pfizer vials but in
282
none of the Moderna vials.
283
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 12
284
Figure 3. The amplification curve for spike (red), ori (blue), and SV40 enhancer-promoter-
285
ori (green) in a single vial of Pfizer (Lot: Fx4343a) from two different wells of the same
286
PCR run.
287
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 13
288
Figure 4. Amplification plot of all Pfizer (A) and Moderna (B) vials showing that spike (red)
289
and ori (blue) amplified similarly for individual vials of Pfizer. In Moderna, inter-vial
290
variability was consistent, but spike amplified earlier than ori (ΔCq~6).
291
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 14
Table 2. Details of the vaccine vials, adverse events (AEs) identified, and qPCR testing results for SARS-CoV-2 spike, ori,
292
and the SV40 promoter-enhancer-ori on all Pfizer-BioNTech and Moderna vials tested. Calculations for Pfizer and Moderna
293
were based on adult doses of 0.30 mL and 0.50 mL, respectively. Moderna is also indicated to be given to children aged
294
6-12 years of age with a dose 0.25 mL making the resultant total ng/dose half of that given to adults. Total ng/dose is
295
adjusted for the length of the amplicon (105 bp ori, 114 bp spike) only representing a fraction of the 7,824 bp Pfizer and
296
6,777 bp Moderna plasmid.
297
Vaccine Information
VAERS Data
Spike
Ori
SV40
Manufacturer
Type
Lot
Number *
Printed
Expiry Date
Total
AES
Total
SAEs
Cq
Total
ng/dose
Total
Copies/dose
Cq
Total
ng/dose
Total
Copies/dose
Cq
Pfizer-BioNTech
Adult Monovalent
FM7380
02/2022
29
15
18.03
2.43
2.07E+10
18.57
3.92
1.86E+11
17.19
Pfizer-BioNTech
Adult Monovalent
FN7934a
08/2022
42
21
18.47
1.79
1.53E+10
18.77
3.43
1.62E+11
16.64
Pfizer-BioNTech
Adult Monovalent
FN7934b
02/2022
18.19
2.18
1.86E+10
18.44
4.27
3.96E+10
16.96
Pfizer-BioNTech
Adult Monovalent
FX4343a
08/2022
1
0
23.53
0.27
2.30E+09
24.71
0.32
2.94E+09
20.64
Pfizer-BioNTech
Adult Monovalent
FX4343b
07/2022
23.83
0.22
1.86E+09
24.87
0.28
2.64E+09
22.59
Pfizer-BioNTech
Adult Bivalent
GK0932a
09/2022
3
0
20.46
2.25
1.92E+10
21.01
3.81
3.54E+10
18.53
Pfizer-BioNTech
Adult Bivalent
GK0932b
09/2022
20.60
2.05
1.75E+10
21.22
3.32
3.08E+10
18.91
Pfizer-BioNTech
Adult Bivalent
GK0932c
09/2022
20.66
1.97
1.68E+10
21.21
3.33
3.09E+10
18.6
Moderna
Child/Adult Monovalent
020E21A
None Stated
5
1
23.66
0.35
3.02E+09
29.47
0.02
1.87E+08
Neg
Moderna
Child/Adult Monovalent
020J21A
30/032022
7
5
23.21
0.48
4.12E+09
30.10
0.01
1.23E+08
Neg
Moderna
Child/Adult Monovalent
033M21Aa
22/06/2022
2
1
23.04
0.54
4.65E+09
29.46
0.02
1.88E+08
Neg
Moderna
Child/Adult Monovalent
033M21Ab
30/07/2022
22.81
0.64
5.44E+09
29.38
0.02
1.99E+08
Neg
Moderna
Child/Adult Monovalent
033M21Ac
30/03/2022
23.59
0.37
3.18E+09
29.87
0.02
1.43E+08
Neg
Moderna
Child/Adult Monovalent
033M21Ad
30/07/2022
23.26
0.47
3.98E+09
29.39
0.02
1.97E+08
Neg
Moderna
Child/Adult Monovalent
055K21A
30/07/2022
2
2
22.94
0.58
4.98E+09
29.58
0.02
1.74E+08
Neg
Moderna
Child/Adult Monovalent
062H21Aa
30/07/2022
9
3
22.52
0.78
6.69E+09
29.21
0.02
2.23E+08
Neg
Moderna
Child/Adult Monovalent
062H21Ab
28/05/2022
22.76
0.66
5.64E+09
29.37
0.02
2.00E+08
Neg
Moderna
Adult Bivalent BA.4/5
AT0709Ba
30/07/2023
0
0
23.68
0.35
2.99E+09
29.30
0.02
2.09E+08
Neg
Moderna
Adult Bivalent BA.4/5
AT0709Bb
30/07/2023
23.56
0.38
3.24E+09
29.25
0.02
2.16E+08
Neg
Moderna
Adult Bivalent BA.4/5
AT0709Bc
30/07/2023
23.63
0.36
3.09E+09
29.34
0.02
2.04E+08
Neg
Moderna
Adult Bivalent BA.4/5
AT0709Bd
30/07/2023
23.80
0.32
2.74E+09
29.44
0.02
1.91E+08
Neg
Moderna
Child/Adult Bivalent BA.1
AS0467Da
02/04/2023
0
0
23.20
0.49
4.17E+09
25.24
0.34
3.20E+09
Neg
Moderna
Child/Adult Bivalent BA.1
AS0467Db
02/04/2023
24.16
0.25
2.14E+09
26.08
0.20
1.82E+09
Neg
Moderna
Child/Adult Bivalent BA.1
AS0467Dc
02/04/2023
23.75
0.33
2.85E+09
25.74
0.25
2.28E+09
Neg
Moderna
Adult Monovalent XBB.1.5
020G23Aa
29/04/2024
0
0
24.42
0.73
6.26E+09
29.42
0.03
3.18E+08
Neg
Moderna
Adult Monovalent XBB.1.5
020G23Ab
29/04/2024
24.46
0.71
6.11E+09
29.87
0.03
2.33E+08
Neg
Moderna
Adult Monovalent XBB.1.5
020G23Ac
29/04/2024
24.53
0.68
5.84E+09
29.74
0.03
2.55E+08
Neg
*Lower case letters at the end of lot numbers indicate different vials of the same lot. SV40 promoter-enhancer-ori
298
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 15
DNA content in none of the Moderna and three Pfizer lots exceeded 1 ng/dose for either
299
spike or plasmid ori. Vaccine in these three vials was diluted 10-fold serially to assess
300
LNP inhibition in qPCR (Figure 5). We observed the expected ~3.3 Cq response after
301
the 1:10 dilution (1:10, 1:100, 1:1000) suggesting that there is some LNP inhibition that
302
could impact the quantitation of DNA at these dilutions (Figure 6). Therefore, the data
303
from the 1:10 dilutions were used for further analysis. This dilution, as well as the fact
304
that some of the doses were designed to be diluted before use, was accounted for in our
305
calculations.
306
307
Figure 5. Comparison of residual DNA content of spike (red) and ori (blue) and the total
308
number of adverse events (orange) reported to VAERS. The FDA and WHO regulatory
309
guideline of 10 ng/dose13 14 for residual DNA is shown by a red dotted line. Vials are
310
sorted in descending order by DNA load of plasmid ori. Lower case letters at the end of
311
lot numbers indicate different vials of the same lot. The total number of AEs was
312
determined per lot and reproduced for each vial in the same lot.
313
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 16
314
Figure 6. qPCR amplification profiles from the serial dilutions (10-fold) of the three lots
315
containing the highest DNA loads (Pfizer lots: A, FN7934a; B, FN7934b; C, FM7380).
316
317
The amount of residual DNA varied substantially between lots (0.28 - 4.27 ng/dose for
318
Pfizer ori, 0.22 - 2.43 ng/dose for Pfizer spike, 0.01 - 0.34 ng/dose for Moderna ori, 0.25-
319
0.78 ng/dose for Moderna spike) when tested by qPCR. Fluorometer based
320
measurements (e.g., Qubit®) of the vaccines show 2,567 ± 618 ng/dose (range: 1,896
321
to 3,720 ng/dose) for Pfizer and 4,280 ± 593 ng/dose (range: 3,270 to 5,100 ng/dose)
322
for Moderna suggesting a high fraction of the DNA is under the size range of the qPCR
323
amplicons.
324
325
We plotted residual DNA values obtained by Qubit fluorometry against those obtained
326
by qPCR (Figure 7). For the Pfizer product, the trend lines for ori and spike estimates
327
both had a positive slope. The graph for the Moderna product differs from that of the
328
Pfizer product with little overlap of values in either axis, with much shallower slopes.
329
Although a detailed view of the Moderna plots suggests a negative slope for the ori
330
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 17
values, this trendline may be influenced by three outlying values. These values were
331
obtained from vials of the Moderna BA.1-Wuhan bivalent vaccines.
332
333
334
Figure 7. Graphical comparison of residual DNA concentration for spike (red) and ori
335
(blue) determined by qPCR and total residual DNA concentration in individual vials as
336
determined by Qubit. In panel A both Pfizer and Moderna data are plotted on the same
337
scale. The Moderna data are enclosed in a red box and displayed separately with an
338
enlarged scale in panel B, to display detail.
339
340
Other than Moderna lots AS0709D, AS0467D and 020G23A, VAERS reports were found
341
for all lots examined in this study (Figure 5). Of the 12 lots examined, the lots with the
342
highest numbers of reports filed to VAERS worldwide were FM7380 and FN7934 with
343
29 and 42 reports, respectively. In the case of lot FM7380, 15 individuals (52%) reported
344
an SAE, whereas for lot FN7934, 1 individual died, 2 individuals reported a disability,
345
and 18 reported being hospitalized with 21 (50%) SAEs. There were 9, 7, 5, 3, 2, and 2
346
reports filed for lots 062H21A, 020J21A, 020E21A, GK0932, 033M21A and 055K21A,
347
respectively. Of these lots, 5/7 (71%) reports for Moderna lot 020J21A involved
348
hospitalization, and there were 1/5 reports of death for Moderna lot 020E21A. In total
349
there were 100 reports of AEs filed worldwide to VAERS for these lots; 48 (48%) of these
350
were SAEs. Most of these AE (n=92) and SAE (n=44) reports originated from outside
351
the USA in similar proportion. Of these 92 AEs, 70 (76%) could be identified as
352
originating in Canada, with another 5 (5.4%) whose origin could not be determined.
353
354
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 18
In an exploratory analysis, we constructed dose-response curves by plotting (Figure 8)
355
the mass of DNA for spike (red) and plasmid ori (blue) found in Pfizer (upper panel) and
356
Moderna (lower panel) vials against the SAE reporting ratio (SRR). The ori and spike
357
curves for the Pfizer product are similar to each other and show a positive dose-response
358
relationship. The corresponding curves for the Moderna lots are shifted leftwards by one
359
to two orders of magnitude. However, the ori and spike curves differ in position and
360
slope.
361
362
Figure 8. Exploratory dose-response analysis comparing the concentration of residual
363
DNA measured by qPCR for spike (red) and plasmid ori (blue) found in Pfizer (A) and
364
Moderna (B) lots plotted against the SRR (reports of SAEs / total number of all adverse
365
events reported to VAERS) for each lot from countries outside of the USA. Residual
366
DNA mass per dose is plotted on a logarithmic scale. Data from the 1:10 dilution were
367
used.
368
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 19
The corresponding plots for residual DNA estimated using fluorometry (Figure 9) yielded
369
curves with a negative slope for both the Pfizer and Moderna products.
370
371
372
Figure 9. Exploratory dose-response analysis comparing the concentration of residual
373
DNA measured by Qubit fluorometry for Pfizer (blue) and Moderna (red) vaccine lots
374
plotted against the SRR (reports of SAEs / total number of all adverse events reported
375
to VAERS) from countries outside of the USA. Residual DNA mass per dose is plotted
376
on a logarithmic scale.
377
378
The Pfizer children’s monovalent (Lot FL8095) described by McKernan et al.4 was
379
sequenced with Oxford Nanopore (ONT) to assess the read length distributions after
380
mapping the reads to the reference sequence of the plasmid in NCBI (Figure 8). The
381
longest read detected in 865 reads was 3.5 kb with read mapping to most of the plasmid
382
backbone (Figure 9).
383
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 20
384
Figure 10. Oxford Nanopore (ONT) read length distributions from 866 reads mapped to
385
the vector sequence (NCBI OR134577.1). Mean = 214 bp. Max = 3.5 kb.
386
387
Figure 11. Longest Oxford Nanopore (ONT) read aligns to the vector region shown in
388
blue. ori and spike primer locations are annotated on the innermost circle. Open reading
389
frames (ORFs) are annotated in gold and green arrows. Kanamycin resistance genes
390
were detected in a very shallow sequencing survey of the vaccine.
391
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 21
Nuclease sensitivity of the Pfizer vaccine was assessed using DNaseI-XT. This DNA
392
nuclease is optimized for IVT reactions rich in RNA:DNA hybrids. This treatment showed
393
<1 Cq offset while a naked DNA control spiked into LNPs was reduced from a Cq of 15
394
to undetectable under the same conditions. This indicates that the DNA present in the
395
vaccines is protected by encapsulation in the LNPs (Figure 10, Figure 11).
396
397
Figure 12. DNase I-XT treatment of Pfizer vaccine demonstrates nuclease resistance
398
of the DNA in the vaccines.
399
400
Figure 13. DNaseI-XT positive control demonstrates the digestion assay eliminates all
401
spiked in DNA under the same conditions used to assess the vaccine nuclease
402
sensitivity.
403
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 22
Discussion
404
Residual DNA was detected in all 27 vaccine vials surveyed. Multiple vials from the same
405
lots produced very similar loads for all targets showing assay reliability, reproducibility,
406
and consistency within the lots. These data involving vaccine vials distributed in Canada
407
are consistent with several non-peer reviewed reports of DNA contamination in modRNA
408
vaccines (McKernan, Buckhaults, Konig).4 15 16
409
410
Moderna had the lowest DNA concentration by qPCR but the highest concentration with
411
Qubit. The Moderna vials had the most consistent levels of DNA between vials
412
suggesting a more robust and standardized manufacturing process. In each vial of the
413
Moderna product, except for lot AS0467D, ori displayed lower loads than spike
414
suggesting a more effective removal of the vector DNA. Possibly, homologous modified
415
RNA may prevent digestion of template DNA by hybridization.17
416
417
The vials with the highest DNA concentration were from two lots of Pfizer monovalent
418
purple top vials with a phosphate buffered saline (PBS) formulation and require dilution
419
before administration. On October 29, 2021, the US FDA authorized a change of
420
formulation to a Tris/sucrose buffer; the grey topped monovalent adult vaccine and an
421
orange topped vaccine for children aged 6-11 years, This change was made to increase
422
stability, to simplify storage requirements and to provide a ready-to-use formulation.18 19
423
These purple-topped Pfizer lots were also associated with the highest number of AEs
424
and SAEs reported in VAERS among all the lots tested. As the actual number of doses
425
administered for each lot is unknown, we used the total number of AEs as a proxy for
426
the number of doses administered as a denominator for the number of SAEs to estimate
427
toxicologic/pharmacologic effect. This uses the same principle used by CDC10 11 in its
428
disproportionality signal analysis (DSA).
429
430
Our exploratory analysis of the relationship between the residual DNA content and SAEs
431
reported to VAERS is preliminary and limited in sample size but warrants confirmation
432
by examining many more lots and vials. A positive dose-response relationship was
433
observed for the Pfizer lots based on qPCR estimation of residual DNA.
434
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 23
Different relationships were observed for Moderna lots for qPCR data as well as for plots
435
based on residual DNA estimated by fluorometry, for both Pfizer and Moderna lots.
436
These observations may reflect differences between the two products such as quantity
437
of DNA, the size distribution of DNA fragments, the composition and sequence of the
438
plasmid vector and composition of lipid nanoparticles. Other differences both between
439
the two products and between different lots of each product may also contribute to our
440
observations. These differences include variations in levels of contaminants or
441
impurities. One major source of impurity is fragmented mRNA for which a number of
442
toxicological mechanisms have been proposed such as its effects on miRNA
443
processes.20 dsRNA is another type of impurity that occurs secondarily to the T7 RNA
444
polymerase promoter. dsRNA can induce pro-inflammatory cytokines21 and has been
445
hypothesized to contribute to immune-inflammatory reactions such as myocarditis.22
446
Lipopolysaccharides in cells from endotoxin can bind both the S1 and S2 subunits of the
447
spike protein which may result in enhanced inflammatory responses.23
448
449
Wider sampling will likely reveal greater detail in terms of event types, such as death, as
450
well as comparisons with other works such as that reported by Schmeling et al.24 who
451
reported a correlation of AEs to various vaccine lot numbers24. None of the presently
452
studied vaccine lots were included in the Schmeling study and more work is needed to
453
understand if and how this DNA contamination is related to AEs.
454
455
While the SV40 enhancer facilitates nuclear localization,6 25 genomic integration of DNA
456
fragments has yet to be demonstrated for the COVID-19 modRNA products.26 However,
457
it is known that DNA contamination could trigger an unwarranted innate immune
458
response and may be prothrombotic, particularly for fragments with high GC content.27
459
dsDNA may also be a significant factor in ischemic diseases including stroke.28 While
460
there appears to be a correlation between high DNA contamination and SAEs more
461
research is needed to expand the sample size and elucidate any potential mechanism
462
at work.
463
464
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 24
It is important to emphasize that because qPCR cannot quantitate molecules smaller
465
than the size of the amplicon (105-114 bp), qPCR underestimates the total DNA in each
466
vaccine. This explains the large differences we have observed in residual DNA levels
467
estimated by qPCR compared with Qubit fluorometry particularly between the Pfizer and
468
Moderna products. The much larger values obtained for the Moderna product suggests
469
that there is a higher fraction of small fragmented residual DNA than in the Pfizer
470
product. This is consistent with a more thorough nuclease digestion step. This illustrates
471
the DNA contamination guidelines recommended by the FDA are highly dependent on
472
the methods used to quantitate the DNA. An alternative hypothesis to explain the high
473
fluorometric measurements is the unknown specificity of the DNA-tropic fluorometric
474
dyes when in use with samples that have high concentrations of N1-methyl-
475
pseudouridine modRNA.
476
477
This fluorometry assessment is of particular interest as fluorometry and UV
478
spectrophotometry were used to quantitate RNA in the Pfizer COVID-19 vaccines, as
479
described in EMA documents3, while qPCR was used to quantitate DNA. This selective
480
use of different methods to quantitate RNA/DNA ratios can lead to vastly different results
481
for the ratio-metric guidelines in place at the EMA.
482
483
This elevated fluorometry quantitation compared to qPCR quantitation is consistent with
484
the ONT read length distributions that also suggest a portion of the DNA may be smaller
485
than the amplicon size. While the ONT sequencer detects molecules shorter than 100
486
bp, the methods for library construction for ONT use a 0.7X Ampure DNA purification
487
step which drastically selects against purifying molecules <150 bp in size. As a result,
488
the read length distributions for ONT reads are biased towards fragments >150 bp and
489
are not a perfect reflection of the smaller fragments that may be present and
490
undercounted by both ONT and qPCR.
491
492
Currently, the US FDA recommends manufacturers of viral vaccines to limit the amount
493
of residual DNA in the final product to below 10 ng/dose for parenteral inoculations and
494
the size of the DNA to below the size of a functional gene, or ~200 base pairs.13 This is
495
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 25
also in keeping with recommendations from the World Health Organization (WHO).14 29
496
Previous residual DNA levels were set by the FDA at 10pg/dose in 1985. A 1986 WHO
497
study group concluded that the risk is negligible up to 100 pg/dose and in 1996 the WHO
498
further increased levels up to 10 ng per dose.14
499
500
The FDA and WHO guidelines for allowable DNA in vaccines are influenced by work
501
published by FDA scientists Sheng-Fowler et al.30 This work focused on host cell
502
genomic DNA contamination and made note of the increased number of molecules
503
present when small viral vectors are the contaminating species. For these high copy per
504
nanogram contaminants, femtograms to attograms of DNA are considered the
505
equivalent of nanograms of cell substrate genomic DNA. Given the short fragment size
506
in the modRNA vaccines, the number of molecules in each dose can reach over 100
507
billion molecules. The residual DNA in these vaccines is high in copy number and rich
508
in promoters, ORFs and nuclear targeting sequences. The FDA and WHO guidelines
509
did not consider packaging of DNA in lipid nanoparticles, likely resulting in longer DNA
510
persistence as well as increased transfection efficiency. Furthermore, the guidelines did
511
not consider cumulative dosing with LNP-based modRNA. In some cases, more than
512
five doses of COVID-19 vaccines have been administered with a dose interval for
513
booster doses sometimes as short as 2 months. Moreover, the risks of cumulative dosing
514
by vaccines targeting other infections but using the same plasmid and LNP-based
515
modRNA platform has not been considered in setting the residual DNA guidelines.
516
517
The FDA guidelines are also written to only quantitate DNA fragments of 200 bp or
518
greater, in part because fragments smaller than this were not considered to be able to
519
produce a functional gene. However, Klinman et al.,31 suggests that fragments as small
520
as 7bp can pose integration risks. Furthermore, the guidelines may also have considered
521
that fragments of naked DNA shorter than 200 bp would be more rapidly hydrolyzed by
522
host nucleases activity than larger molecules.32 This accelerated destruction cannot be
523
assumed of the vaccines due to the DNA being encapsulated and protected by the LNPs.
524
525
526
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 26
Klinman et al.,31 also observe that “in evaluating the potential harm of plasmid
527
integration, it should be noted that the risk of introducing plasmids with strong regulatory
528
regions into the host genome far exceeds that associated with random point mutations.”
529
530
Finally, the guidelines do not consider if the residual DNA contains nuclear targeting
531
sequences and mammalian promoters that exist in the Pfizer vaccine.26 Vacik et al.
532
demonstrated that the SV40 enhancer present in the Pfizer vector is a potent nuclear
533
targeting sequence showing promise for gene therapy.25
534
535
Conclusion
536
These data demonstrate the presence of billions to hundreds of billions of DNA
537
molecules per dose in the modRNA COVID-19 products tested. Using fluorometry, all
538
products tested exceeded the guidelines for residual DNA set by the FDA and WHO of
539
10 ng/dose by 188 – 509-fold. However, qPCR detected residual DNA content in all
540
products tested were below these guidelines emphasizing the importance of
541
methodological clarity and consistency when interpreting quantitative guidelines. The Cq
542
scores for the most recent XBB.1.5 Moderna vaccine suggest that DNA residues have
543
not been reduced from previous vaccine versions.
544
545
The preliminary evidence of a dose-response effect of residual DNA measured with
546
qPCR and SAEs warrants confirmation and further investigation. Our findings extend
547
existing concerns about vaccine safety and call into question the relevance of guidelines
548
conceived before the introduction of efficient transfection using LNPs. With several
549
obvious limitations, we urge that our work is replicated under forensic conditions and
550
that guidelines be revised to account for highly efficient DNA transfection and cumulative
551
dosing.
552
553
This work highlights the need for regulators and industry to adhere to the precautionary
554
principle, and provide sufficient and transparent evidence that products are safe and
555
effective, and disclose the details of their composition and method of manufacture.
556
557
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 27
Data Availability
558
Fastq file for the mapped ONT sequencing data:
559
https://mega.nz/file/UZhkiTBQ#8vjDK5JV5N5Dj2On34B6zdRObEKGBy3ZC7w8q2t9U
560
Vc
561
562
Acknowledgements
563
We thank all the pharmacists who donated to our endeavors.
564
565
Author Contributions
566
DJ Speicher: sample management, study design, qPCR, data analysis, manuscript
567
preparation
568
J Rose: VAERS analysis, manuscript preparation
569
LM Gutschi: data analysis, manuscript preparations
570
D Wiseman: data analysis, manuscript preparations
571
K McKernan: qPCR assay design, DNAseI and ONT experiments, manuscript
572
preparation
573
574
Conflict of Interest Statement
575
Kevin McKernan is employed by Medicinal Genomics and provided qPCR reagents free
576
of charge. The other authors declare that there are no conflicts of interest.
577
578
Revision History
579
2023-10-19 – version 1.0
580
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 28
References
581
1. Guetzkow JA. Letters to the editor: Effect of mRNA Vaccine Manufacturing Processes
582
on Efficacy and Safety Still an Open Question. BMJ 2023;378:o1731.
583
2. Tinari S. The EMA covid-19 data leak, and what it tells us about mRNA instability.
584
BMJ 2021;372:n627. doi: 10.1136/bmj.n627 [published Online First: 20210310]
585
3. EMEA/H/C/005735/RR. Rapporteur Rolling Review Critial Assessment Report:
586
Quality Aspects. 2020 [Available from: https://covidvaccinereactions.com/ema-
587
pfizer-leak/.
588
4. McKernan K, Helbert Y, Kane LT, et al. Sequencing of bivalent Moderna and Pfizer
589
mRNA vaccines reveals nanogram to microgram quantities of expression vector
590
dsDNA per dose. OSFio 2023 doi: 10.31219/osf.io/b9t7m
591
5. Shimabukuro TT, Nguyen M, Martin D, et al. Safety monitoring in the Vaccine Adverse
592
Event Reporting System (VAERS). Vaccine 2015;33(36):4398-405. doi:
593
10.1016/j.vaccine.2015.07.035 [published Online First: 20150722]
594
6. Dean DA, Dean BS, Muller S, et al. Sequence requirements for plasmid nuclear
595
import. Exp Cell Res 1999;253(2):713-22. doi: 10.1006/excr.1999.4716
596
7. New England Biolabs. NEBioCalculator 2023 [Available from:
597
https://nebiocalculator.neb.com/#!/dsdnaamt2023.
598
8. R: The R Project for Statistical Computing [program]: R Foundation for Statistical
599
Computing, 2010.
600
9. Funk PR, Yogurtcu ON, Forshee RA, et al. Benefit-risk assessment of COVID-19
601
vaccine, mRNA (Comirnaty) for age 16-29 years. Vaccine 2022;40(19):2781-89.
602
doi: 10.1016/j.vaccine.2022.03.030 [published Online First: 20220328]
603
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 29
10. VAERS Team. Vaccine Adverse Event Reporting System (VAERS) Standard
604
Operating Procedures for COVID-19 (as of 29 January 2021) 2021 [Available from:
605
https://www.cdc.gov/vaccinesafety/pdf/VAERS-v2-SOP.pdf.
606
11. Evans SJ, Waller PC, Davis S. Use of proportional reporting ratios (PRRs) for signal
607
generation from spontaneous adverse drug reaction reports. Pharmacoepidemiol
608
Drug Saf 2001;10(6):483-6. doi: 10.1002/pds.677
609
12. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler
610
transform. Bioinformatics 2009;25(14):1754-60. doi:
611
10.1093/bioinformatics/btp324 [published Online First: 20090518]
612
13. Food and Drug Administration. Chemistry, Manufacturing, and Control (CMC)
613
Information for Human Gene Therapy Investigational New Drug Applications
614
(INDs): U.S. Department of Health and Human Services; 2020 [Available from:
615
https://www.fda.gov/media/113760/download.
616
14. Shin J, Wood D, Robertson J, et al. WHO informal consultation on the application of
617
molecular methods to assure the quality, safety and efficacy of vaccines, Geneva,
618
Switzerland, 7-8 April 2005. Biologicals 2007;35(1):63-71. doi:
619
10.1016/j.biologicals.2005.12.005 [published Online First: 20060220]
620
15. South Carolina Senate. SC Senate Hearing - USC Professor Dr. Phillip Buckhaults
621
2023 [updated 2023-09-13. Available from:
622
https://www.youtube.com/watch?v=IEWHhrHiiTY.
623
16. Koning B. Letter regarding the evidence presented to you of massive DNA
624
contamination in BioNTech's COVID-19 mRNA vaccine.
625
https://www.genimpfstoffe.com/wp-content/uploads/2023/09/Verblindet-Fax-
626
BMG-Lauterbach-16.-Sept-2023-1.pdf, 2023.
627
17. Sutton DH, Conn GL, Brown T, et al. The dependence of DNase I activity on the
628
conformation of oligodeoxynucleotides. Biochem J 1997;321 ( Pt 2)(Pt 2):481-6.
629
doi: 10.1042/bj3210481
630
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 30
18. U.S. Food and Drug Administration. Emergency Use Authorization (EUA) for an
631
Unapproved Product Review Memorandum 2021 [Available from:
632
https://www.fda.gov/media/153947/download.
633
19. Food and Drug Administration. Letter of Authorization 2023 [Available from:
634
https://www.fda.gov/media/150386/download.
635
20. Stati G, Amerio P, Nubile M, et al. Concern about the Effectiveness of mRNA
636
Vaccination Technology and Its Long-Term Safety: Potential Interference on
637
miRNA Machinery. Int J Mol Sci 2023;24(2) doi: 10.3390/ijms24021404 [published
638
Online First: 20230111]
639
21. Nelson J, Sorensen EW, Mintri S, et al. Impact of mRNA chemistry and
640
manufacturing process on innate immune activation. Sci Adv
641
2020;6(26):eaaz6893. doi: 10.1126/sciadv.aaz6893 [published Online First:
642
20200624]
643
22. Milano G, Gal J, Creisson A, et al. Myocarditis and COVID-19 mRNA vaccines: a
644
mechanistic hypothesis involving dsRNA. Future Virol 2021 doi: 10.2217/fvl-2021-
645
0280 [published Online First: 20211206]
646
23. Samsudin F, Raghuvamsi P, Petruk G, et al. SARS-CoV-2 spike protein as a
647
bacterial lipopolysaccharide delivery system in an overzealous inflammatory
648
cascade. J Mol Cell Biol 2023;14(9) doi: 10.1093/jmcb/mjac058
649
24. Schmeling M, Manniche V, Hansen PR. Batch-dependent safety of the BNT162b2
650
mRNA COVID-19 vaccine. Eur J Clin Invest 2023;53(8):e13998. doi:
651
10.1111/eci.13998 [published Online First: 20230409]
652
25. Vacik J, Dean BS, Zimmer WE, et al. Cell-specific nuclear import of plasmid DNA.
653
Gene Ther 1999;6(6):1006-14. doi: 10.1038/sj.gt.3300924
654
Speicher DJ et al, DNA fragments detected in COVID-19 vaccines in Canada. 31
26. Lim S, Yocum RR, Silver PA, et al. High spontaneous integration rates of end-
655
modified linear DNAs upon mammalian cell transfection. Sci Rep 2023;13(1):6835.
656
doi: 10.1038/s41598-023-33862-0 [published Online First: 20230426]
657
27. Gaitzsch E, Czermak T, Ribeiro A, et al. Double-stranded DNA induces a
658
prothrombotic phenotype in the vascular endothelium. Sci Rep 2017;7(1):1112.
659
doi: 10.1038/s41598-017-01148-x [published Online First: 20170425]
660
28. Ma X, Xin D, She R, et al. Novel insight into cGAS-STING pathway in ischemic
661
stroke: from pre- to post-disease. Frontiers in Immunology 2023;14
662
29. World Health Organization. Annex 3: Requirements for the use of animal cell cultures
663
as substrates for the manufacture of biologic medicinal products and the
664
characterization of cell banks. Replacement of Annex 1 of WHO Technical Report
665
Series, No. 878. 2013 [Available from: https://cdn.who.int/media/docs/default-
666
source/biologicals/documents/trs_978_annex_3.pdf?sfvrsn=fe61af77_3&downloa
667
d=true.
668
30. Sheng-Fowler L, Lewis AM, Jr., Peden K. Issues associated with residual cell-
669
substrate DNA in viral vaccines. Biologicals 2009;37(3):190-5. doi:
670
10.1016/j.biologicals.2009.02.015 [published Online First: 20090314]
671
31. Klinman DM, Klaschik S, Tross D, et al. FDA guidance on prophylactic DNA
672
vaccines: analysis and recommendations. Vaccine 2010;28(16):2801-5. doi:
673
10.1016/j.vaccine.2009.11.025 [published Online First: 20091124]
674
32. André M, Reghin S, Boussard E, et al. Universal real-time PCR assay for quantitation
675
and size evaluation of residual cell DNA in human viral vaccines. Biologicals
676
2016;44(3):139-49. doi: 10.1016/j.biologicals.2016.03.002 [published Online First:
677
20160328]
678
679
