Impact of COVID-19 disease and COVID-19 vaccination on maternal or fetal inflammatory response, placental pathology, and perinatal outcomes
COVID-19和疫苗接种对母体/胎儿炎症反应、胎盘病理学和围产期结果的影响
10.1016/j.ajog.2022.05.049
2022-05-28, Article
Abstract & Authors:展开
Abstract:收起
OBJECTIVE: COVID-19 disease, even mild, is associated with a host inflammatory response1, which has not
been well delineated in pregnancy. COVID-19 disease is associated with preterm birth2,
preeclampsia2, and placental vascular pathology3. While COVID-19 vaccination does result in
mild viral-type symptoms associated with a heighted immune response, it is not known whether
this immune response results in inflammatory cytokine pathway activation. Our objective is to
evaluate the impact of COVID-19 disease and vaccination on the maternal-fetal unit through
study of inflammatory cytokine panel at delivery, placental pathology, and perinatal outcomes.
STUDY DESIGN:
This is a retrospective cohort study of pregnant patients who delivered at Thomas
Jefferson University Hospital from March 2020-July 2021 and consented to an ongoing delivery
sample biorepository. With our convenience sample of ~300 we anticipated at least 15% rate of
COVID disease or vaccination, or N≥50 in each group. The cohort was categorized as 1)
Control- no COVID-19 disease or vaccination history; 2) COVID-19 disease (PCR) during
pregnancy, and 3) COVID-19 mRNA vaccination during pregnancy (patient report). Those with
both COVID-19 disease and vaccination were excluded. Placental histopathology was sent for
standard clinical criteria. Cytokine paneling completed on those with maternal or cordblood
samples available and without a diagnosis of clinical triple-I. Cytokine analysis was done using
the Mesoscale Diagnostics (MSD) platform using the 10-plex human pro-inflammatory panel kit
(IL1β, IL8, IL2, IL4, IL6, IL10, IL12p70, IL13, IFNγ, TNFα) and read on a QuickPlex SQ120
(Mesoscale, Gaithersburg, MD).
RESULTS: 306 were included, Control (N=198), COVID+ (N=59) and COVID vaccine (N=49).
There were significant differences in age, BMI, and race between groups (Supplemental table 1).
Mean latency to delivery was similar between COVID vaccine and COVID positive groups
(89.5±46.6 days vs 71.0 ± 78.6 days respectively, p=0.14), although more in the COVID-vaccine
group had latency >30 days from delivery compared to COVID disease (91.7% vs 60.7%,
p<0.001). Of those with COVID, 54 (91.5%) were mild, none were severe.
Maternal cytokine analysis included N=142 control, N=58 COVID disease, and N=42
COVID vaccine. COVID-19 vaccination had lower median TNFα concentrations than COVIDdisease
or control (p=0.001), while COVID-disease had higher median IL-6 compared to control
(p=0.04) (Figure 1a, supplemental Figure 1). IL-6 and TNFα levels were not correlated with
latency from exposure to delivery (> or <30 days) among those with COVID-19 disease (Pearson
correlation coefficient r=0.10, p=0.48 and r=0, p>0.99 respectively) or among those with
COVID-19 vaccination (r=0.12, p=0.46 and r=0.14, p=0.40 respectively).
Among cordblood samples, COVID-19 disease, but not vaccination, was associated with
elevated cord blood IL-10 compared to control (p=0.04), although this did not remain significant
after Bonferroni correction (p=0.14) (Figure 1a, supplemental Figure 2).
Correlation matrices were generated to assess differences in the cytokine environment
between groups. Significant correlation between IL12/IL13/IL-2 was seen in control maternal
samples and IL12/IL13/IL4 in COVID-vaccine maternal serum samples. These relationships
suggest a balanced Th1/Th2 profile without a pro-inflammatory tendency. Samples from
COVID-19 positive maternal serum samples shows a strong correlation between
IL10/TNFα/IL1β/IL8 indicating a pro-inflammatory milieu (Figure 1b). Cytokine milieu in cord
blood samples were similar across cohorts (Figure 1c). There was a positive correlation betweencertain maternal and cord blood Th1 and Th2 markers (IL-12, IL-13, IL-2, IL-4, and TNFα)
(Supplemental Figure 3).
Using multivariable logistic regression we evaluated association of cytokines with the
outcomes of preeclampsia, preterm birth, and placental maternal vascular malperfusion (MVM).
No maternal cytokines were associated with preeclampsia or MVM. TNFα was positively
associated with preterm birth (B= 1.58 (1.15-2.15, p=0.004) (Supplemental table 4a-c). Similar
findings were noted in the subgroup of those with documented COVID-19 infection
(Supplemental Table 5a-c); TNFα remained significantly associated with preterm birth risk
(B=3.77 (1.34-10.61), p=0.01)).
COVID-19 disease, but not vaccination, was associated with a significantly higher rate of
placental MVM (55.8% vs 35.8%, aOR 2.34 (1.16-4.73)) and preterm birth (15.3% vs 6.6%,
aOR 4.54 (1.61-12.78)) compared to control (Table 1). Among those with COVID-19 during
pregnancy and placental pathology available (N=52), 19 reported aspirin use. The rate of
placental MVM was lower with aspirin, although not statistically significant (42.5% vs 63.6%,
p=0.13). In contrast, among N=95 control, the rate of placental MVM was higher in the aspirin
group (48.5% vs 29.0%, p=0.06), as expected based on baseline risk factors prompting aspirin
use. There was no correlation between placental MVM and COVID-19 latency (greater or less
than 30 days) (r= -0.10, p=0.51).
CONCLUSION: COVID-19 disease, but not vaccination, in pregnancy is associated with increased
inflammatory cytokines, placental vascular pathology, and preterm delivery. This inflammatory
signature, specifically TNFα, was correlated with preterm birth risk but not placental vascular
pathology, suggesting potentially divergent pathways related to these downstream consequences.19 vaccination, despite mild viral-like symptoms, did not elicit the same pathological
effects, providing both pathophysiological and clinical evidence of safety of mRNA vaccines in
pregnancy. Compared to a smaller study4, we did not identify a fetal inflammatory signature
associated with COVID-19. It is likely there was undocumented COVID-19 disease exposure in
the control or vaccine groups3, although it would not be possible to determine if this exposure
were prior to or during pregnancy, and if anything this would lead to Type 1 error. Further
research is needed to elaborate on the pathophysiological basis for adverse perinatal outcomes
following maternal COVID-19, timing of infection and risk of these outcomes, and how
therapies, such as aspirin, may mitigate these downstream perinatal effects5.
First Authors:
Rupsa C Boelig
Correspondence Authors:
Rupsa C Boelig
All Authors:
Rupsa C Boelig,Zubair H Aghai,Sidhartha Chaudhury,Adina S Kazan,Joanna S Y Chan,Elke Bergmann-Leitner
