ORIGINAL RESEARCH • CARDIAC IMAGING

Assessment of Myocardial 18F-FDG Uptake at

PET/CT in Asymptomatic SARS-CoV-2–vaccinated and
Nonvaccinated Patients
Takehiro Nakahara, MD, PhD • Yu Iwabuchi, MD, PhD • Raita Miyazawa, MD • Kai Tonda, MD •
Tohru Shiga, MD, PhD • H. William Strauss, MD • Charalambos Antoniades, MD, PhD •
Jagat Narula, MD, PhD • Masahiro Jinzaki, MD, PhD
From the Department of Radiology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan (T.N., Y.I., R.M., K.T., T.S., M.J.);
Advanced Clinical Research Center, Fukushima Medical University, Fukushima, Japan (T.S.); Department of Cardiology, Icahn School of Medicine at Mount Sinai, New
York, NY (H.W.S.); Division of Cardiovascular Medicine, Radcliffe Department of Medicine, University of Oxford Radcliffe Hospital, Oxford, UK (C.A.); and Department
of Medicine and Cardiology, McGovern Medical School, Houston, Tex (J.N.). Received March 31, 2023; revision requested May 25; revision received July 29; accepted
August 2. Address correspondence to T.N. (email: [email protected]).
Conflicts of interest are listed at the end of this article.
See also the editorial by Bluemke in this issue.

Radiology 2023; 308(3):e230743 • https://doi.org/10.1148/radiol.230743 • Content code:

Background: Patients who developed myocarditis after SARS-CoV-2 vaccination show abnormalities on cardiac MRI scans.
However, whether myocardial changes occur in asymptomatic individuals after vaccination is not well established.

Purpose: To assess myocardial fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake on PET/CT images in asymptomatic patients
vaccinated against SARS-CoV-2 compared with nonvaccinated patients.

Materials and Methods: This retrospective study included patients who underwent 18F-FDG PET/CT for indications unrelated to
myocarditis during the period before (November 1, 2020, to February 16, 2021) and after (February 17, 2021, to March 31, 2022)
SARS-CoV-2 vaccines were available. Myocardial and axillary 18F-FDG uptake were quantitatively assessed using maximum stan-
dardized uptake value (SUVmax). The SUVmax in all patients and in patients stratified by sex (male or female), age (<40 years, 41–60
years, >60 years), and interval between vaccination and PET/CT were compared using the Mann-Whitney U test or the Kruskal-
Wallis test with post ad hoc Dwass-Steel-Critchlow-Fligner multiple comparison analysis.

Results: The study included 303 nonvaccinated patients (mean age, 52.9 years ± 14.9 [SD]; 157 female, 146 male) and 700
vaccinated patients (mean age, 56.8 years ± 13.7; 344 female, 356 male). Vaccinated patients had overall higher myocardial
18
F-FDG uptake compared with nonvaccinated patients (median SUVmax, 4.8 g/mL [IQR, 3.0–8.5 g/mL] vs 3.3 g/mL [IQR,
2.5–6.2 g/mL]; P < .001). Myocardial SUVmax was higher in vaccinated patients regardless of patient sex (median range, 4.7–4.9
g/mL [IQR, 2.9–8.6 g/mL]) or age (median range, 4.7–5.6 g/mL [IQR, 2.9–8.6 g/mL]) compared with corresponding nonvac-
cinated groups (sex: median range, 3.2–3.9 g/mL [IQR, 2.4–7.2 g/mL]; age: median range, 3.3–3.3 g/mL [IQR, 2.3–6.1 g/
mL]; P < .001 to P = .015). Furthermore, increased myocardial 18F-FDG uptake was observed in patients imaged 1–30, 31–60,
61–120, or 121–180 days after their second vaccination (median SUVmax range, 4.6–5.1 g/mL [IQR, 2.9–8.6 g/mL]) (P < .001
to P = .001), and increased ipsilateral axillary uptake was observed in patients imaged 1–30, 31–60, and 61–120 days after their
second vaccination (median SUVmax range, 1.5–2.0 g/mL [IQR, 1.2–3.4 g/mL]) compared with the nonvaccinated patients (P <
.001 to P < .001).

Conclusion: When compared with nonvaccinated patients, asymptomatic patients who received their second vaccination 1–180 days
prior to imaging showed increased myocardial 18F-FDG uptake on PET/CT scans.
© RSNA, 2023

Supplemental material is available for this article.

W hile vaccines to prevent SARS-CoV-2 infection

have demonstrated effectiveness in reducing mor-
bidity and mortality related to respiratory complications
MRI scans and 18F-FDG uptake on PET scans in patients
suspected of having myocarditis (12).
A recent cardiac MRI study used late gadolinium en-
(1,2), infrequent but important side effects associated hancement and T2 intensity and reported myocardial in-
with vaccination have also been reported. One such rare jury from the SARS-CoV-2 vaccine was similar to that from
side effect that the mRNA vaccines have been linked to myocarditis due to COVID-19, while severity was less (13).
is myocarditis (3–7). Similarly, an 18F-FDG PET/MRI study showed myocardial
Cardiac MRI (4,7,8) and fluorine 18 (18F) fluorodeoxy- inflammation after COVID-19 illness (14), but it is not
glucose (FDG) PET/CT imaging (9–11) have been rou- known whether 18F-FDG uptake would occur in asymp-
tinely used in the noninvasive diagnosis of myocardial in- tomatic individuals after SARS-CoV-2 vaccination.
flammation of diverse origin, including viral myocarditis, The purpose of the current study was to semiquantita-
cardiac sarcoidosis, and cancer therapy–related cardiac dys- tively and quantitatively assess myocardial 18F-FDG uptake
function. Good agreement has been reported between late on PET/CT scans in asymptomatic patients vaccinated
gadolinium enhancement or T2 hyperintensity on cardiac against SARS-CoV-2 and asymptomatic nonvaccinated
This copy is for personal use only. To order copies, contact [email protected]
Assessment of Myocardial 18F-FDG Uptake

18
F FDG to scanning]) were excluded from analysis. Patients
Abbreviations who had a history of infection with SARS-CoV-2 or who had
FDG = fluorodeoxyglucose, SUV = standardized uptake value, received a third dose were excluded from the current study.
SUVmax = maximum SUV
If patients had undergone multiple examinations during the
Summary study period, the most recent study was used for the main
Asymptomatic patients who underwent PET/CT 1–180 days after analysis (Fig 1).
their second SARS-CoV-2 vaccination showed increased myocardial
fluorine 18 fluorodeoxyglucose uptake on images compared with
nonvaccinated patients, but patients imaged more than 180 days after PET/CT Procedure
vaccination did not. All patients were routinely instructed to skip a meal, and vital
Key Results signs and blood glucose level were measured prior to 18F-FDG
■ In a retrospective study of 700 patients vaccinated against SARS- injection using a blood glucose meter (Medisafe FIT Pro II;
CoV-2 and 303 nonvaccinated patients who underwent PET/CT for Terumo) and lancing device (Medisafe Finetouch II; Terumo).
indications other than myocarditis, patients who received their second
Approximately 60 minutes after intravenous injection of 4.0
vaccination 1–180 days before imaging showed higher myocardial
fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake (median MBq per kilogram of body weight 18F-FDG, whole-body
standardized uptake value [SUVmax] range, 4.6–5.1 g/mL [IQR, PET/CT images were acquired with integrated PET/CT sys-
2.9–8.6 g/mL]) than nonvaccinated patients (median SUVmax, 3.3 g/ tems (Biograph mCT or Biograph Vision 600; Siemens Medi-
mL [IQR, 2.5–6.2 g/mL]; range, P < .001 to P = .001).
cal Solutions). Low-dose CT (100-kVp tube voltage, 50-mAs
■ Myocardial 18F-FDG uptake was higher in vaccinated patients tube current, 0.5 second per rotation, 2-mm section thickness)
regardless of sex or patient age compared with corresponding
nonvaccinated groups. was performed for attenuation correction and anatomic co-
registration. No iodinated contrast material was administered.
PET images were acquired from the vertex to the feet in three-
patients who underwent imaging for indications unrelated to dimensional mode for 2 minutes per bed position without re-
myocardial inflammation. spiratory or cardiac gating.

Materials and Methods Evaluation of PET/CT Images

Because of the retrospective nature of this study, the need for All PET/CT images were transferred to a workstation and recon-
written informed consent was waived by our local institutional structed into coronal, axial, and sagittal planes with dedicated
review board. software (AW Server on Universal Viewer; GE HealthCare)
by two observers. One observer (T.N., 20 years of expertise in
Study Sample cardiology) assessed all PET/CT images and assessed PET/CT
This retrospective study took advantage of a large consecutive images from the consecutive initial 71 patients 3 months later
repository of 18F-FDG PET/CT studies that were obtained at to assess intraobserver viability. The other observer (Y.I., 15 years
our institution between November 2020 and March 2022 in of expertise in nuclear medicine) assessed PET/CT images from
adult patients to evaluate various malignancies or other unre- the same consecutive initial 71 patients to assess interobserver
lated indications, including comprehensive medical checkup. viability. Image evaluations were conducted independently, and
The study included one group of patients who had received one observers were blinded to clinical data and previous PET/CT
or two doses of the vaccine for SARS-CoV-2 with clear vaccina- images. For visual analysis of myocardial FDG activity, a scale
tion documentation from February 17, 2021, (the start of the of standardized uptake value (SUV) was set from 0.0 g/mL to
vaccination program in Japan) to March 31, 2022, and a second 6.0 g/mL. Myocardial visual scores were assessed using the fol-
group of patients who did not receive the SARS-CoV-2 vaccine lowing scale: 0, minimal uptake; 1, mostly minimal or mild up-
during this period or during the period before vaccination was take; 2, mostly intense or moderate uptake; and 3, homogeneous
available (November 1, 2020, to February 16, 2021). uptake (17) (Fig S1). For quantitative analysis, a volume of inter-
Patients who had blood glucose levels greater than 100 mg/dL est was set that included the whole heart and axillary nodes in
(5.55 mmol/L) at the time of 18F-FDG injection or who had the ipsilateral side, and a maximum SUV (SUVmax [in grams per
fasted for less than 12 hours (15) were also excluded. Patients milliliter]) (10) was measured. A 10-mm volume of interest was
were also excluded if they had pre-existing diseases or conditions also set to measure SUVmax in the liver and spleen.
that could artifactually influence the myocardial FDG uptake.
Specifically, patients with hematologic diseases, such as lym- Statistical Analyses
phoma and leukemia, cardiac sarcoidosis, and thyroid disease Continuous data were tested for normality with the Kol-
(16); those who had undergone cardiac surgery, chemotherapy mogorov-Smirnov test. Nonnormally distributed continuous
likely to result in cardiac dysfunction, or chest irradiation within data are presented as median and IQR, and normally distributed
the past 6 months; and patients currently undergoing anti- continuous data are presented as mean ± SD. Continuous data
inflammatory therapy, were all excluded. Because body move- were compared using the Mann-Whitney U test between the
ments and scan delay can affect the analysis, patients with two groups or the Kruskal-Wallis test with post ad hoc Dwass-
hard body movement or delayed scan timing (over 10 minutes Steel-Critchlow-Fligner multiple comparison analysis. For blood
[ie, over 70 minutes have passed from intravenous injection of pressure, analysis of covariance was used, adjusting for age.

2radiology.rsna.org ■ Radiology: Volume 308: Number 3—September 2023

 Nakahara et al

who have been diagnosed with cancer or who

showed three points of myocardial visual score
were excluded from the analysis.
To assess whether myocardial 18F-FDG up-
take differed based on the interval between vac-
cination and imaging, patients were divided
into different interval groups, and axillary and
myocardial SUVmax was compared between each
group using the Kruskal-Wallis test with post
ad hoc Dwass-Steel-Critchlow-Fligner multiple
comparison analysis. A previous study showed
effectiveness of the SARS-CoV-2 vaccine against
COVID-19 was high during the 1st month after
the second dose, declined after 4 months, and
was effective until 6 months (18). Thus, we di-
vided the patients into 2-month groups until 6
months (60, 120, and 180 days after vaccination)
and divided them by 1 month (30 days after vac-
cination). To divide 30 days also seems reason-
able because previous studies reported relatively
high risk of myocarditis within 30 days after the
second dose of mRNA vaccines (3–7).
Subjects with unknown vaccination dates or
with ChAdOx1 nCoV-19 (Astra Zeneca) and mis-
cellaneous vaccines were excluded when patients
were stratified by interval or type of vaccine due
to the small sample size. In a subanalysis of pa-
tients with more than one scan, 18F-FDG uptake
was compared across multiple scans using the Wil-
coxon signed rank test.
Two-sided P < .05 was considered indicative
of a significant difference. Statistical analyses were
performed by an observer (T.N.) using SAS soft-
ware (version 9.4; SAS Institute).
In the initial impression from 200 consecu-
Figure 1: Flow diagram shows patient exclusion criteria. Of the cumulative total 9478 patients tive patients, the average myocardial SUVmax was
who underwent fluorine 18 fluorodeoxyglucose PET/CT, 1003 patients matched the study criteria, 6.2 g/mL in 139 vaccinated patients and 4.8 g/mL
including 700 patients from the period during which SARS-CoV-2 vaccines were available (February in 61 nonvaccinated patients at 8 weeks; therefore,
17, 2021, to March 31, 2022) and 303 patients from the period before SARS-CoV-2 vaccines were effect size was estimated to be 0.36, and allocation
available (November 1, 2020 to February 16, 2021, n = 125) and 178 patients from the period after
(February 17, 2021, to March 31, 2022). A total of 25 patients (four nonvaccinated, 21 vaccinated)
ratio was estimated to be 2.3. We determined that
underwent two examinations during the study period, but only the second study was used. a target of 280 nonvaccinated and 644 vaccinated
A subanalysis including these patients was also performed using the first and second scans. adjudicated primary outcomes would provide a
power of 99% at a two-sided α level of .01.

Categorical data are presented as proportions and percentages Results

and were compared with the χ2 test or Fisher exact probability
test, as appropriate. Patient Characteristics
Intra- and interobserver variability of myocardial uptake In total, 9478 patients with available PET examinations were
scores were assessed with the Cohen κ coefficient. The strength initially considered for inclusion. Patients younger than 20 years
of agreement for κ values is as follows: κ < 0.20 = poor agree- (n = 19), those with a blood glucose level higher than 100 mg/dL
ment; κ = 0.21–0.40, fair agreement; κ = 0.41–0.60, moder- (5.55 mmol/L) (n = 6201), those with insufficient fasting at the
ate agreement; κ = 0.61–0.80, good agreement; and κ = 0.81– time of FDG injection (n = 1137), those with pre-existing diseases
1.00, excellent agreement. Agreement between SUVmax values (sarcoidosis, n = 42; thyroid disease, n = 64; hematologic
in the axilla, myocardium, liver, and spleen was assessed us- malignancy, n = 408) or treatments in the past 6 months (surgery,
ing Bland-Altman analysis, and linear correlation was assessed n = 69; chemotherapy, n = 205; irradiation, n = 21; anti-
using Spearman rank correlation. In a subanalysis of patients inflammatory therapy, n = 32) that could artifactually influ-
without cancer or patients with homogeneous uptake, patients ence myocardial FDG uptake, those in whom an inappropriate

Radiology: Volume 308: Number 3—September 2023 ■ radiology.rsna.org 3

Assessment of Myocardial 18F-FDG Uptake

When compared with nonvaccinated patients, vaccinated

Patient Characteristics
patients had a higher myocardial 18F-FDG uptake visual score
No Vaccine Vaccine
(median, 2 [IQR, 0–3] vs 1 [IQR, 0–2]; P < .001) (Fig 3A)
Characteristic (n = 303) (n = 700) P Value and SUVmax (median, 4.8 g/mL [IQR, 3.0–8.5 g/mL] vs 3.3 g/mL
[IQR, 2.5–6.2 g/mL]; P < .001) (Fig 3B), which remained af-
No malignancies 150 (49.5) 372 (53.1) …
ter age-adjustment for both measures (P < .001). In patients
Sex
Female 157 (51.8) 344 (49) .44
without cancer, 372 vaccinated individuals also showed a higher
Male 146 (48.2) 356 (50.9) … median myocardial FDG uptake visual score (median, 2 [IQR,
Age (y) 52.9 ± 14.9 56.8 ± 13.7 <.001* 0–3]) and SUVmax (median, 4.8 g/mL [IQR, 3.2–8.3 g/mL])
Height (cm) 163.1 ± 9.0 164.0 ± 8.8 .12 compared with 150 nonvaccinated individuals (median visual
Weight (kg) 60.4 ± 15.2 61.5 ± 13.2 .08 score, 1 [IQR, 0–2]; median SUVmax, 3.3 g/mL [IQR, 2.6–
Blood pressure 6.3 g/mL]; P < .001 for both).
Systolic (mmHg) 121.5 ± 17.1 124.3 ± 18.0 .014† When only patients with myocardial visual scores less than
Diastolic (mmHg) 76.8 ± 11.2 78.4 ± 12.2 .057 3 were analyzed, the vaccinated group (n = 479) demonstrated
Hypertension 36 (11.9) 108 (15.4) .14 higher myocardial FDG uptake visual scores (median, 1 [IQR,
Dyslipidemia 15 (5.0) 75 (10.7) .003* 0–2]) and SUVmax (median, 3.6 g/mL [IQR, 2.7–5.1 g/mL])
Diabetes 1 (0.003) 11 (0.02) .12 compared with the nonvaccinated group (n = 248; median visual
Hyperuricemia 5 (0.02) 15 (0.02) .81 score, 0 [IQR, 0–1]; median SUVmax, 3.0 g/mL [IQR, 2.4–4.1 g/
Note.—Categorical variables are presented as numbers of mL]; P < .001 for both).
patients, with percentages in parentheses. Continuous variables Myocardial SUVmax remained higher in the vaccinated group,
are presented as means ± SDs. The χ2 test and Fisher exact even after dividing by liver SUVmax (median, 2.1 g/mL [IQR,
probability test were used to compare categorical variables, and 1.3–3.6 g/mL]) or splenic SUVmax (median, 2.5 g/mL [IQR,
the Mann-Whitney U test was used to compare continuous
variables between the two groups.
1.6–4.4 g/mL]), when compared with the nonvaccinated group
* P < .05.
(median when divided by liver SUVmax, 1.5 g/mL [IQR, 1.1–2.7
†
Significance lost after adjusting for age.
g/mL]; P < .001; median when divided by splenic SUVmax, 1.9 g/
mL [IQR, 1.3–3.3 g/mL]; P < .001). The vaccinated group also
showed higher FDG uptake in the liver (median SUVmax, 2.3
scan was performed (n = 7), those with no vaccine information g/mL [IQR, 2.1–2.5 g/mL]) and spleen (median SUVmax, 1.9
(n = 165), those with previous SARS-CoV-2 infection (n = 13), g/mL [IQR, 1.7–2.1 g/mL]) compared with the nonvaccinated
and those who received a third dose of the vaccine (n = 65) (Fig 1) group (median liver SUVmax, 2.2 g/mL [IQR: 2.0–2.4 g/mL]; P
were excluded. Ultimately, the study included 1003 patients; 303 < .001; median spleen SUVmax, 1.9 g/mL [IQR: 1.7–2.0 g/mL];
were not vaccinated (157 female, 146 male) and 700 were vac- P = .007).
cinated (344 female, 356 male; 40 patients had one vaccine dose,
660 patients had two vaccine doses) at the time of PET/CT im- Myocardial 18F-FDG Uptake in Patients with Vaccination
aging (mean incubation time, 60.2 minutes ± 0.9). Patient char- Side Effects
acteristics are presented in the Table. Vaccinated patients were After vaccination, 254 of 700 (36.3%) patients reported a fever,
older (mean age, 56.8 years ± 13.7) than nonvaccinated patients and 458 of 700 (65.4%) reported a sore arm, but no patients
(mean age, 52.9 years ± 14.9; P < .001) and more frequently reported chest pain. The myocardial visual score was higher in
had dyslipidemia (nonvaccinated, 5.0% [15 of 303]; vaccinated, patients who reported a sore arm (median score, 1 [IQR, 0–3])
10.7% [75 of 700]; P = .003). Systolic blood pressure was also compared with those who did not (median score, 2 [IQR, 1–3];
higher in the vaccinated group (mean, 124.3 mmHg ± 18.0) P = .032), but no difference was observed in myocardial SUVmax
than in the nonvaccinated group (mean, 121.5 mmHg ± 17.1; between the two groups (median, 4.6 g/mL [IQR, 2.8–8.2 g/mL]
P = .014) but not after adjustment for age (P = .31). In the vac- vs 4.9 g/mL [IQR, 3.2–8.6 g/mL]; P = .09). Additionally, no dif-
cinated group, 372 of 700 (53.1%) patients did not have can- ference in visual score (median, 2 [IQR, 0–3] vs 2 [IQR, 0–3];
cer, whereas in the nonvaccinated group, 150 of 303 (49.5%) P = .40) or myocardial SUVmax (median, 4.8 g/mL [IQR, 3.1–8.1
patients did not have cancer. Details on the types of malignan- g/mL] vs 4.8 g/mL [IQR, 2.9–9.3 g/mL]; P = .36) was observed
cies patients had and the therapies they underwent more than between patients who developed a fever and those who did not.
6 months prior to imaging are reported in Tables S1 and S2,
respectively. Assessment of Myocardial 18F-FDG Uptake in Patients
Stratified by Interval between Vaccination and PET/CT
Assessment of Myocardial 18F-FDG Uptake Based on Patients were divided into seven groups based on the interval
Vaccination Status between vaccination and imaging: (a) nonvaccinated, (b) imag-
Myocardial 18F-FDG uptake score and quantification of ing after the first dose, and imaging (c) 30 days or less, (d) 31–60
18
F-FDG uptake in the axilla, myocardium, liver, and spleen showed days, (e) 61–120 days, (f) 121–180 days, and (g) more than
excellent intra- and interobserver agreement (Appendix S1, 180 days after the second dose. Patients with an unclear date of
Fig S2). Representative PET/CT images with myocardial vaccination were excluded from this analysis (n = 8). The me-
18
F-FDG uptake are shown in Figure 2 and Figures S3 and S4. dian duration from the first vaccine dose to PET imaging was

4radiology.rsna.org ■ Radiology: Volume 308: Number 3—September 2023

 Nakahara et al

Figure 2: Representative whole-body and myocardial fluorine 18 (18F) fluorodeoxyglucose (FDG) PET images (top row) coronal and axial PET images (middle row),
and axial color-blending PET/CT fusion images (bottom row) in patients vaccinated against SARS-CoV-2 and nonvaccinated patients. (A) Images in a 43-year-old man
who underwent 18F-FDG PET/CT for comprehensive medical checkup during the period before SARS-CoV-2 vaccines were available. The patient had a myocardial
score of 2 and a myocardial maximum standardized uptake value (SUVmax) of 2.7 g/mL. Axillary, liver, and spleen SUVmax were 0.6, 2.8, and 2.1 g/mL, respectively.
(B) Images in an 80-year-old man with pancreatic cancer who underwent PET/CT before SARS-CoV-2 vaccines were available. The patient had a myocardial score of 0
and a myocardial SUVmax of 2.2 g/mL. Axillary, liver, and spleen SUVmax were 1.1, 2.2, and 1.5 g/mL, respectively. (C) Images in a 38-year-old man who underwent PET/
CT for comprehensive medical checkup 29 days after he received the first dose of the BNT16b2 vaccine in the left arm. High uptake of 18F-FDG in the left axilla (arrow) and
myocardium were observed. The patient had a myocardial score of 3 and a myocardial SUVmax of 14.6 g/mL. Axillary, liver, and spleen SUVmax were 5.0, 2.0, and 2.1 g/mL,
respectively. (D) Images in a 72-year-old man who underwent PET/CT for comprehensive medical checkup 139 days after he received the second dose of the mRNA-1273
vaccine in the left arm. High uptake of 18F-FDG in the left axilla (arrow) and myocardium were observed. The patient had a myocardial score of 2 and a myocardial SUVmax
of 5.9 g/mL. Axillary, liver, and spleen SUVmax were 2.7, 2.6, and 2.1 g/mL, respectively.

13 days (range, 6–21 days), and the median duration from the but this was not observed in patients who underwent imag-
second dose to PET imaging was 88 days (range, 41–135 days). ing more than 120 days after their second vaccination (median
Patients who underwent imaging 1–180 days after receiving their SUVmax range, 1.1–1.2 g/mL [IQR, 0.9–1.5 g/mL]; P = .20 to
second vaccination had higher myocardial FDG uptake (median P = .99) (Fig 4A, Table S3).
SUVmax range, 4.6–5.1 g/mL [range of IQRs, 2.9–8.6 g/mL])
than nonvaccinated patients (median SUVmax, 3.1 g/mL [IQR, Myocardial 18F-FDG Uptake in Patients Stratified by Sex
2.5–6.2 g/mL]; P < .001 to P = .001), but patients imaged more and Age
than 180 days after their second dose did not (median SUVmax, When patients were stratified by sex, myocardial 18F-FDG uptake
4.5 g/mL [IQR, 2.7–9.3 g/mL]; P = .15) (Fig 4B). Furthermore, was higher in vaccinated male patients (median SUVmax, 4.9 g/mL
higher axillary FDG uptake was observed in patients who un- [IQR, 3.3–8.6 g/mL]) than in nonvaccinated male patients
derwent imaging 1–120 days after receiving their second vac- (median SUVmax, 3.9 g/mL [IQR, 2.7–7.2 g/mL]; P = .004)
cination (median SUVmax range, 1.5–2.0 g/mL [range of IQRs, and higher in vaccinated female patients (median SUVmax, 4.7
1.2–3.4 g/mL]) compared with nonvaccinated patients (median g/mL [IQR, 2.9–8.2 g/mL]) than in nonvaccinated female
SUVmax, 1.2 g/mL [IQR, 1.0–1.4 g/mL]; P < .001 to P < .001), patients (median SUVmax, 3.2 g/mL [IQR, 2.4–5.1 g/mL];

Radiology: Volume 308: Number 3—September 2023 ■ radiology.rsna.org 5

Assessment of Myocardial 18F-FDG Uptake

Figure 3: Qualitative and quantitative assessment of myocardial fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake in vaccinated and nonvaccinated patients. (A) Bar
plot shows the number of patients who received each myocardial 18F-FDG uptake visual score (range, 0–3) stratified by vaccination status (nonvaccinated, n = 303;
vaccinated, n = 700). Myocardial 18F-FDG uptake visual scores were higher in the vaccinated group than in the nonvaccinated group (Mann-Whitney U test, P < .001).
(B) Boxplot shows myocardial 18F-FDG uptake measured by maximum standardized uptake value (SUVmax) in nonvaccinated (n = 303) and vaccinated (n = 700) patients.
Myocardial SUVmax was higher in the vaccinated group (median, 4.8 g/mL [IQR, 3.0–8.5 g/mL]) than in the nonvaccinated group (median, 3.3 g/mL [IQR, 2.5–6.2 g/mL];
P < .001). Horizontal bars in the boxplot represent median SUVmax, and whiskers represent interquartile range. The diamond in the box represents the average. Mann-
Whitney U test was used to compare median SUVmax between groups.

P < .001) (Fig 5A). The axillary uptake was also higher in vac- both vaccinated groups (P < .001 for both), with no difference
cinated male (median SUVmax, 1.4 g/mL [IQR, 1.1–1.8 g/mL]) in 18F-FDG uptake observed between BNT162b2 mRNA (me-
and female (median SUVmax, 1.5 g/mL [IQR, 1.1–1.9 g/mL]) dian SUVmax, 4.7 g/mL [IQR, 2.9–8.4 g/mL]) and mRNA-1273
patients than in nonvaccinated patients of either sex (median (median SUVmax, 5.1 g/mL [IQR, 3.4–8.7 g/mL]; P = .39) vac-
SUVmax in male patients, 1.2 g/mL [IQR, 1.0–1.5 g/mL]; cine types. Axillary SUVmax was higher in both the BNT162b2
P < .001; median SUVmax in female patients, 1.2 g/mL [IQR, mRNA group (median, 1.4 g/mL [IQR, 1.1–1.8 g/mL]) and the
1.0–1.4 g/mL]; P < .001) (Fig S5A). mRNA-1273 group (median, 1.5 g/mL [IQR, 1.1–2.0 g/mL])
Patients were also stratified into three age groups: those than in the nonvaccinated group (median, 1.2 g/mL [IQR, 1.0–
less than 40 years of age, those aged 41–60 years, and those 1.4 g/mL]; P < .001 for both) (Fig S6A, S6B).
aged more than 60 years. For each age group, the 18F-FDG up-
take of the axilla and myocardium were higher in vaccinated Myocardial 18F-FDG Uptake in a Subset of Patients with
(median SUVmax range, 4.7–5.6 g/mL [IQR, 2.9–8.6 g/mL]) Multiple PET/CT Studies
than in nonvaccinated (median SUVmax range, 3.3–3.3 g/mL A total of 25 patients had more than one PET/CT study avail-
[IQR, 2.3–6.1 g/mL]; P < .001 to P = .015) patients (Fig 5B). able. Among them, 16 patients who had not undergone che-
However, no difference in myocardial or axillary FDG uptake motherapy underwent PET/CT both before vaccination and
was observed between vaccinated (median SUVmax range, within 180 days after their second vaccination (median inter-
1.4–1.6 g/mL [IQR, 1.1–1.9 g/mL]) and nonvaccinated (median val, 87.5 days [IQR, 56.5–104.5 days]; range 16–158 days).
SUVmax range, 1.1–1.3 g/mL [IQR, 0.7–1.6 g/mL]; P < .001 to Compared with FDG uptake on PET/CT scans obtained be-
P < .001) patients in each age group (Fig S5B). fore vaccination, both axillary and myocardial 18F-FDG uptake
were higher on scans obtained after vaccination (difference in
Myocardial 18F-FDG Uptake in Patients Stratified by Type axillary SUVmax, 0.2 g/mL [IQR, 0.1–0.7 g/mL]; P = .028)
of Vaccine (difference in myocardial SUVmax, 1.0 g/mL [IQR, 0.2–2.8 g/mL];
Of the vaccinated patients, the majority (543 of 700 [77.6%]) P = .037) (Fig 6).
received BNT162b2 mRNA (Pfizer-BioNTech), while 147 of
700 (21.0%) received mRNA-1273 (Moderna). Patients who re- Discussion
ceived ChAdOx1 nCoV-19 (AstraZeneca) (one of 700 [0.1%]) Although patients who developed myocarditis after SARS-
or miscellaneous types (nine of 700 [1.3%]) were excluded from CoV-2 vaccination show abnormalities on cardiac MRI scans,
analysis because of the small sample size. As compared with the whether myocardial changes occur in asymptomatic individuals
unvaccinated group (median myocardial SUVmax, 3.3 g/mL after SARS-CoV-2 vaccination is not well established. It was re-
[IQR, 2.5–6.2 g/mL]), the myocardial SUVmax was higher in ported that fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake

6radiology.rsna.org ■ Radiology: Volume 308: Number 3—September 2023

 Nakahara et al

Figure 4: Boxplots show fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake in the (A) axillary and (B) myocardium of patients stratified by the
interval between SARS-CoV-2 vaccination and PET/CT imaging. (A) Compared with the unvaccinated group (dose 0, median maximum standard-
ized uptake value (SUVmax), 1.2 g/mL [IQR, 1.0–1.4 g/mL]), the axillary SUVmax was higher in patients imaged after their first dose (median, 1.6 g/mL
[IQR, 1.3–3.2 g/mL]; P < .001). Patients imaged within 30 days (median, 2.0 g/mL [IQR, 1.6–3.4 g/mL]), 31–60 days (median, 1.7 g/mL
[IQR, 1.5–1.9 g/mL]), and 61–120 days (median, 1.5 g/mL [IQR, 1.2–1.7 g/mL]) after they received their second dose of the vaccine also showed
increased axillary SUVmax compared with the unvaccinated group (P <.001 to P < .001). There was no difference observed in axillary SUVmax between
unvaccinated patients and patients imaged 121–180 days (median, 1.2 g/mL [IQR, 1.0–1.5 g/mL]; P = .99) or more than 180 days (median, 1.1 g/
mL [IQR, 0.9–1.3 g/mL]; P = .20) after their second dose. (B) Boxplot shows myocardial SUVmax for nonvaccinated (dose 0) and vaccinated groups.
The myocardial SUVmax was higher in patients imaged after their first dose (median, 6.2 g/mL [IQR, 3.8–8.8 g/mL]; P = .004) and in patients imaged
1–30 days (median, 5.1 g/mL [IQR, 3.2–8.6 g/mL]), 31–60 days (median, 4.8 g/mL [IQR, 3.0–7.7 g/mL]), 61–120 days (median, 4.6 g/mL [IQR,
3.2–8.5 g/mL]), and 121–180 days (median, 5.1 g/mL [IQR, 2.9–8.2 g/mL]) after their second dose compared with the unvaccinated group (me-
dian, 3.3 g/mL [IQR, 2.5–6.2 g/mL]; P < .001 to P < .001). There was no difference observed in myocardial SUVmax between unvaccinated patients
and patients imaged more than 180 days after their second dose (median, 4.5 g/mL [IQR, 2.7–9.3 g/mL]; P = .15). For both boxplots, horizontal bars
represent median SUVmax and whiskers represent interquartile range. The diamond in the box represents the average. Kruskal-Wallis test with post ad
hoc Dwass-Steel-Critchlow-Fligner multiple comparison analysis was used to compare median SUVmax between groups.

on PET/CT scans correlated with late gadolinium enhancement except for the vaccinated group including individuals imaged
or T2 intensity on cardiac MRI scans in patients with CO- more than 180 days after their second vaccination (median
VID-19 myocarditis. The aim of this study was to investigate SUVmax, 4.5 g/mL [IQR, 2.7–9.3 g/mL]; P = .15). No differ-
myocardial 18F-FDG uptake on PET/CT scans in asymptomatic ence in myocardial or axillary 18F-FDG uptake was observed
patients vaccinated against SARS-CoV-2 compared with uptake between patients who received the BNT162b2 mRNA vac-
in nonvaccinated patients. cine and those who received the mRNA-1273 vaccine. In 16
In this observational study of patients who underwent patients with more than one PET/CT study available, myo-
PET/CT during comprehensive medical check-ups or to eval- cardial and axillary 18F-FDG uptake were higher on PET/CT
uate malignancies, patients who had received a SARS-CoV-2 scans obtained after vaccination than those obtained before
mRNA-based vaccine showed increased myocardial 18F-FDG vaccination.
uptake on scans compared with nonvaccinated patients (median Although infrequent, incidences of myocarditis have been re-
visual score, 2 [IQR, 0–3] vs 1 [IQR, 0–2]; P < .001; median ported after SARS-CoV-2 vaccination (3–7,19–21) in patients
SUVmax, 4.75 g/mL [IQR, 3.0–8.5 g/mL] vs 3.3 g/mL [IQR, younger than 40 years (6,19,21), in both male (4,5,21,22) and
2.5–6.2 g/mL]; P < .001). This increase in myocardial 18F-FDG female patients (6), and in patients who received the mRNA-
uptake in vaccinated patients was also observed in subgroup 1273 vaccine (6,19) and those who received the BNT162b2
analyses that excluded individuals with cancer or homogeneous mRNA vaccine (20). In our study, no differences in myocardial
myocardial uptake. When patients were divided into groups 18
F-FDG uptake were observed in vaccinated patients when
based on the interval between vaccination and imaging, myo- stratified by age, sex, or vaccine type.
cardial 18F-FDG uptake was higher in all vaccinated groups Several studies have also reported that myocarditis incidents
(median SUVmax range, 4.6–5.1 g/mL [range of IQRs, 2.9– occurred no more than 28 days after patients had received their
8.6 g/mL]) compared with the nonvaccinated group (median second vaccination (3–7,19,21). In our study, patients who under-
SUVmax, 3.1 g/mL [IQR, 2.5–6.2 g/mL]; P < .001 to P = .001) went imaging 1–180 days after their second vaccination showed

Radiology: Volume 308: Number 3—September 2023 ■ radiology.rsna.org 7

Assessment of Myocardial 18F-FDG Uptake

Figure 5: Boxplots show myocardial fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake as measured with maximum standardized uptake
value (SUVmax) in nonvaccinated (-) and vaccinated (+) patients stratified by (A) sex and (B) age. (A) For both sexes, myocardial 18F-FDG
uptake was higher in the vaccinated group (male median SUVmax, 4.9 g/mL [IQR, 3.3–8.6 g/mL]; female median SUV max, 4.7 g/mL [IQR,
2.9–8.2 g/mL]) than in the nonvaccinated group (male median SUVmax, 3.9 g/mL [IQR, 2.7–7.2 g/mL]; P < .001; female median SUVmax, 3.2 g/
mL [IQR, 2.4–5.1 g/mL]; P < .001). (B) For each patient age group assessed, myocardial SUVmax was higher in the vaccinated group (patients
aged <40 years: median SUVmax, 5.6 g/mL [IQR, 3.1–8.5 g/mL]; patients aged 41–60 years: median SUVmax, 4.7 g/mL [IQR, 3.0–8.6 g/mL];
patients aged >60 years: median SUVmax, 4.7 g/mL [IQR, 2.9–8.3 g/mL]) than in the nonvaccinated group (patients aged <40 years: median
SUVmax, 3.3 g/mL [IQR, 2.3–6.1 g/mL]; patients aged 41–60 years: median SUVmax, 3.3 g/mL [IQR, 2.7–6.3 g/mL]; patients aged
>60 years: median SUVmax, 3.3 g/mL [IQR, 2.4–5.5 g/mL]; P < .001 to P < .001). For vaccinated patients, no differences in myocardial SUVmax
were observed between age groups. For both boxplots, horizontal bars represent median SUVmax and whiskers represent interquartile range. The
diamond in the box represents the average. Kruskal-Wallis test with post ad hoc Dwass-Steel-Critchlow-Fligner multiple comparison analysis was
used to compare median SUVmax values between groups.

elevated myocardial 18F-FDG uptake on PET/CT scans compared participants who had fasted for less than 12 hours. This
with nonvaccinated patients, but patients imaged more than potentially led to physiologic uptake and affected the result,
180 days after vaccination did not. A recent cardiac MRI study although it was statistically significant under the same prepara-
reported a similar pattern of myocardial injury between SARS- tion conditions. Third, myocardial 18F-FDG uptake in scans
CoV-2 vaccine–associated myocardial inflammation and other that are not specifically performed to assess cardiac inflamma-
causes of myocardial inflammation but found that vaccine-related tion and that are influenced by many factors (age, sex, insulin
myocardial abnormalities were less severe (13). Thus, even though resistance, diet, etc) are subject to inaccuracies.
vaccinated patients in this study showed elevated myocardial 18F- In conclusion, in a set of patients who underwent PET/
FDG uptake on PET/CT scans up to 180 days after vaccination, CT for indications other than myocardial inflammation, those
this could result from relatively minor inflammation and may not who had received a SARS-CoV-2 vaccine showed increased
represent severe myocardial abnormalities. myocardial fluorine 18 (18F) fluorodeoxyglucose (FDG) uptake
Previous studies have shown that increased 18F-FDG uptake on images up to 180 days after their second vaccination com-
in the axillary lymph nodes of vaccinated patients can persist for pared with patients imaged before SARS-CoV-2 vaccination
2–3 weeks (23–25). Data from the current study suggest this may was available. Vaccinated patients showed higher myocardial
persist for longer, as patients who underwent imaging 1–120 days 18
F-FDG uptake on PET/CT scans compared with nonvacci-
after their second vaccination had higher axillary lymph node nated patients, regardless of sex, age, or type of mRNA vac-
18
F-FDG uptake compared with nonvaccinated patients. When cine received. A prospective study would be needed to validate
compared with cardiac MRI (8), PET/CT can provide informa- the findings of this study, including comparisons with cardiac
tion about inflammation for the whole body, and in the current enzyme levels, cardiac function, and non-mRNA vaccination.
study, 18F-FDG uptake in the liver and spleen was also found to
Acknowledgments: Thanks to Dr Suketaka Momoshima, Dr Shigeo Okuda, Dr
be higher in the vaccinated group versus the nonvaccinated group. Keiichi Narita, and Dr Masashi Tamura for their support to prepare institutional
There were several limitations of this study. First, this was review board documents.
a retrospective study from a single hospital; thus, our findings
may lack generalizability. Second, we did not prepare partici- Author contributions: Guarantors of integrity of entire study, T.N., T.S., H.W.S.;
pants to obviate myocardial glucose uptake, and we excluded study concepts/study design or data acquisition or data analysis/interpretation, all

8radiology.rsna.org ■ Radiology: Volume 308: Number 3—September 2023

 Nakahara et al

Figure 6: Representative whole-body and myocardial

fluorine 18 (18F) fluorodeoxyglucose (FDG) PET/CT images
(coronal and axial PET images, axial color-blending PET/CT
images) in patients who underwent PET/CT both before SARS-
CoV-2 vaccination and within 180 days after their second vaccination
and who were included in a subanalysis. (A) Images in a 54-year-
old woman with vaginal cancer who underwent two PET/CT
examinations, the second of which occurred 107 days after she
received her second dose of the mRNA-1273 vaccine. After vacci-
nation, the myocardial 18F-FDG uptake visual score increased from
1 to 3, and myocardial maximum standardized uptake value
(SUVmax) increased from 4.8 g/mL to 8.2 g/mL on PET/CT scans.
(B) Images in a 67-year-old man with lung cancer who under-
went two PET/CT examinations, the second of which occurred
72 days after he received the second dose of the BNT16b2 vac-
cine. After vaccination, the myocardial 18F-FDG uptake visual score
increased from 1 to 3, and myocardial SUVmax increased from 4.8
g/mL to 18.0 g/mL on PET/CT scans. (C) Images in a 56-year-
old man who underwent two PET/CT examinations for compre-
hensive medical checkup 260 days apart before SARS-CoV-2
vaccines were available. Myocardial score and SUVmax of the first
and second scans, respectively, were 0 and 0 and 1.4 and 2.0
g/mL.

Radiology: Volume 308: Number 3—September 2023 ■ radiology.rsna.org 9

Assessment of Myocardial 18F-FDG Uptake

authors; manuscript drafting or manuscript revision for important intellectual con- 12. Nensa F, Kloth J, Tezgah E, et al. Feasibility of FDG-PET in myocar-
tent, all authors; approval of final version of submitted manuscript, all authors; ditis: Comparison to CMR using integrated PET/MRI. J Nucl Cardiol
agrees to ensure any questions related to the work are appropriately resolved, all 2018;25(3):785–794.
authors; literature research, T.N., T.S., H.W.S., M.J.; clinical studies, T.N., Y.I., 13. Fronza M, Thavendiranathan P, Chan V, et al. Myocardial Injury Pat-
R.M., K.T.; statistical analysis, T.N., R.M.; and manuscript editing, T.N., Y.I., T.S., tern at MRI in COVID-19 Vaccine-Associated Myocarditis. Radiology
H.W.S., C.A., J.N., M.J. 2022;304(3):553–562.
14. Hanneman K, Houbois C, Schoffel A, et al. Combined Cardiac Fluorodeoxy-
Disclosures of conflicts of interest: T.N. Grants from Nihon Medi-Physics and glucose-Positron Emission Tomography/Magnetic Resonance Imaging
Bayer Yakuhin. Y.I. No relevant relationships. R.M. No relevant relationships. K.T. Assessment of Myocardial Injury in Patients Who Recently Recovered
No relevant relationships. T.S. No relevant relationships. H.W.S. No relevant rela- From COVID-19. JAMA Cardiol 2022;7(3):298–308.
tionships. C.A. Founder, shareholder and director of Caristo Diagnostics. J.N. No 15. Yoshinaga K, Miyagawa M, Kiso K, Ishida Y. Japanese Guidelines for Cardiac
relevant relationships. M.J. Grants from Nihon Medi-Physics and Bayer Yakuhin. Sarcoidosis. Ann Nucl Cardiol 2017;3(1):121–124.
16. Biondi B. Endocrine Disorders and Cardiovascular Disease. In: Libby P, Bonow
RO, Mann DL, et al, eds. Braunwald’s heart disease: a textbook of cardio-
References vascular medicine. 12th ed. Philadelphia, Pa: Elsevier, 2022; 1971–1808.
1. Polack FP, Thomas SJ, Kitchin N, et al. Safety and Efficacy of the BNT162b2 17. Williams G, Kolodny GM. Suppression of myocardial 18F-FDG up-
mRNA Covid-19 Vaccine. N Engl J Med 2020;383(27):2603–2615. take by preparing patients with a high-fat, low-carbohydrate diet. AJR
2. Baden LR, El Sahly HM, Essink B, et al. Efficacy and Safety of the mRNA- Am J Roentgenol 2008;190(2):W151–W156.
1273 SARS-CoV-2 Vaccine. N Engl J Med 2021;384(5):403–416. 18. Tartof SY, Slezak JM, Fischer H, et al. Effectiveness of mRNA BNT162b2
3. Verma AK, Lavine KJ, Lin CY. Myocarditis after Covid-19 mRNA Vac- COVID-19 vaccine up to 6 months in a large integrated health system in the
cination. N Engl J Med 2021;385(14):1332–1334. USA: a retrospective cohort study. Lancet 2021;398(10309):1407–1416.
4. Mevorach D, Anis E, Cedar N, et al. Myocarditis after BNT162b2 19. Patone M, Mei XW, Handunnetthi L, et al. Risks of myocarditis, peri-
mRNA Vaccine against Covid-19 in Israel. N Engl J Med 2021; carditis, and cardiac arrhythmias associated with COVID-19 vaccination or
385(23):2140–2149. SARS-CoV-2 infection. Nat Med 2022;28(2):410–422.
5. Witberg G, Barda N, Hoss S, et al. Myocarditis after Covid-19 Vac- 20. Al-Ali D, Elshafeey A, Mushannen M, et al. Cardiovascular and haema-
cination in a Large Health Care Organization. N Engl J Med 2021; tological events post COVID-19 vaccination: A systematic review. J Cell
385(23):2132–2139. Mol Med 2022;26(3):636–653.
6. Husby A, Hansen JV, Fosbøl E, et al. SARS-CoV-2 vaccination and 21. Karlstad Ø, Hovi P, Husby A, et al. SARS-CoV-2 Vaccination and
myocarditis or myopericarditis: population based cohort study. BMJ Myocarditis in a Nordic Cohort Study of 23 Million Residents. JAMA
2021;375:e068665. Cardiol 2022;7(6):600–612.
7. Truong DT, Dionne A, Muniz JC, et al. Clinically Suspected Myocarditis 22. Montgomery J, Ryan M, Engler R, et al. Myocarditis Following Immuni-
Temporally Related to COVID-19 Vaccination in Adolescents and Young zation With mRNA COVID-19 Vaccines in Members of the US Military.
Adults: Suspected Myocarditis After COVID-19 Vaccination. Circulation JAMA Cardiol 2021;6(10):1202–1206.
2022;145(5):345–356. 23. Kubota K, Saginoya T, Ishiwata K, Nakasato T, Munechika H. [18F]
8. Law YM, Lal AK, Chen S, et al. Diagnosis and Management of Myocarditis in FDG uptake in axillary lymph nodes and deltoid muscle after CO-
Children: A Scientific Statement From the American Heart Association. VID-19 mRNA vaccination: a cohort study to determine incidence
Circulation 2021;144(6):e123–e135 [Published correction appears in and contributing factors using a multivariate analysis. Ann Nucl Med
Circulation 2021;144(6):e149.]. 2022;36(4):340–350.
9. Hillner BE, Siegel BA, Liu D, et al. Impact of positron emission tomography/ 24. Advani P, Chumsri S, Pai T, Li Z, Sharma A, Parent E. Temporal metabolic
computed tomography and positron emission tomography (PET) alone response to mRNA COVID-19 vaccinations in oncology patients. Ann
on expected management of patients with cancer: initial results from the Nucl Med 2021;35(11):1264–1269.
National Oncologic PET Registry. J Clin Oncol 2008;26(13):2155–2161. 25. Cohen D, Krauthammer SH, Wolf I, Even-Sapir E. Hypermeta-
10. Chareonthaitawee P, Beanlands RS, Chen W, et al. Joint SNMMI- bolic lymphadenopathy following administration of BNT162b2
ASNC expert consensus document on the role of 18F-FDG PET/CT mRNA Covid-19 vaccine: incidence assessed by [18F]FDG PET-CT
in cardiac sarcoid detection and therapy monitoring. J Nucl Cardiol and relevance to study interpretation. Eur J Nucl Med Mol Imaging
2017;24(5):1741–1758. 2021;48(6):1854–1863.
11. Erba PA, Sollini M, Lazzeri E, Mariani G. FDG-PET in cardiac infec-
tions. Semin Nucl Med 2013;43(5):377–395.

10radiology.rsna.org ■ Radiology: Volume 308: Number 3—September 2023