Cardiotocographic features in COVID-19 infected pregnant women

Introduction

Novel coronavirus 2019 (COVID-19) is an emerging public health problem caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Since its first identification in Wuhan in December 2019, it has been increasing rapidly at an accelerating rate with over 100 million individuals infected [1]. Previous coronavirus epidemics; the severe acute respiratory syndrome coronavirus (SARS-CoV) and the Middle East respiratory syndrome coronavirus (MERS-CoV), have shown that pregnant women and their fetuses are particularly susceptible to poor outcomes such as maternal mortality and morbidity, and perinatal death [2]. There is currently not sufficient knowledge about pregnant women and their complications with COVID-19 in the literature.

Similar to nonpregnant women, fever, cough, myalgia are the most common symptoms of COVID-19 in pregnant women [3, 4]. Although the limited data we have suggests lack of evidence for higher maternal or fetal risks and vertical transmission, fetal complications including preterm labor, fetal growth restriction, fetal distress, and fetal demise have been reported [3, 5, 6].

Cardiotocograph (CTG) is used to record the fetal heart rate (FHR) and determine fetal well-being in order to detect the signs of intrapartum hypoxia [7]. CTG has four features to analyze; baseline FHR, variability, accelerations, and decelerations. These parameters show the activity of fetal somatic and autonomic nervous systems and the oxygenation of the fetal myocardium and brain. In the presence of utero-placental insufficiency, due to hypoxia the fetus switches its metabolism from aerobic to anaerobic and reduces the myocardial workload, and this is reflected in the CTG trace as deceleration [8]. Maternal inflammatory conditions as well as maternal pyrexia can affect fetus and result as baseline tachycardia and decelerations in the CTG trace [9].

In recent studies, it has been shown that SARS-CoV-2 causes an excessively increased response in proinflammatory cytokines (e.g., interleukin-6 [IL-6], interleukin-1β [IL-1β], tumor necrosis factor-α [TNF-α]) and chemokines (e.g., monocyte chemoattractant protein-1 [MCP-1/CCL2]), and hyperactivation of immune cells, causing infiltration of inflammatory cells in the lungs and hypercytokinemia with an exaggerated maternal innate immune response. This condition, called “cytokine storm” is associated with the severity of COVID-19 [10], [11], [12], [13]. Not virus replication itself but cytokine storm leads to acute respiratory distress syndrome and multiorgan failure [10, 11]. Triad of maternal hypoxia seconder to the ARDS, cytokine storm, and hypercoagulability is associated with the risk of placental intervillous thrombosis and infarction [14, 15].

Despite the lack of evidence about COVID-19 vertical transmission to fetus [16], it is likely that fetus would have a reactive response secondary to maternal inflammatory changes and pyrexia, resulting FHR changes reflected in the CTG trace. Maternal inflammatory and hypercoagulable state might be seen as decelerations, absence of variability, and bradycardia on the CTG secondary to uteroplacental insufficiency and placental or umbilical venous thrombosis. Maternal fever as well as maternal inflammatory state, can reduce the utero-placental oxygen transfer and lead to irritation of uterine myometrium which results with increased contractility [17]. In rare cases, severe maternal hypoxia can result as a sinusoidal pattern on the CTG [18].

Since we think that COVID-19 affects FHR as in inflammatory states, in this study, for the first time we aimed to evaluate the CTG traces of 224 pregnant women infected with COVID-19 and analyze whether changes in the CTG traces are related to severity of COVID-19.

Materials and methods

Results

CTG characteristics

The distribution of Apgar 1st min, Apgar 5th min, birth weight and lowest oxygen saturation were given in the Table 3. It was observed that 25.0% (56/224) of patients had minimal or absence of variability (Figure 1), while 4.0% (9/224) had ZigZag pattern (exaggerated variability) [23] (Figure 2). A total of 13.4% (30/224) had no acceleration (Figure 3), and 25.0% (56/224) showed types of decelerations (Figures 3 and 4) A total of 5.4% (12/224) of the patients had tachycardia (Figure 5), while 1.3% (3/224) had bradycardia. 66.5% (149/224) of the patients had uterine contractility, 44.6% (100/224) patients gave birth with C/S, and 90.9% (140/224) did not receive induction. About 22.8% (51/224) of the patients were in labor and 31.2% (70/224) of the patients did not give birth during our study period.

Table 3:

CTG characteristics of the sample.

Variable (n=224)	Mean ± SDa	Median [min–max]	n		%
Variability
Absence of variability			8		3.6
Minimal variability			48		21.4
Normal variability			159		70.9
Marked variabilityb			9		4.0
Tachycardia/bradycardia
No			209		93.3
Tachycardia			12		5.4
Bradycardia			3		1.3
Acceleration
No			30		13.4
Yes			194		86.6
Deceleration
No			168		75.0
Early			5		2.2
Variable			29		12.9
Late			19		8.5
Prolonged			3		1.4
Uterine contractility
No			62		27.7
Yes			149		66.5
Tachysystole			9		4.0
Saw-toothc			4		1.8
Route of delivery
Vaginal delivery			54		24.1
Cesarean delivery			100		44.6
Ongoing pregnancy			70		31.2
Birth induction
Yes			14		9.1
No			140		90.9
Cesarean indications
Previous cesarean delivery			35		35
Preterm labor			11		11
Fetal distress			31		31
Cephalopelvic disproportion			8		8
Induction failure			7		7
Macrosomia			1		1
Malpresentation			3		3
Maternal health condition			4		4
Apgar 1st min	7.52 ± 0.97	8.0 [1.0–9.0]
Apgar 5th min	9.07 ± 0.83	9.0 [6.0–10.0]
Birth weight, g	3,087.79 ± 608.24	3,180.0 [879.0–4,130.0]
Minimum O2 saturation	96.23 ± 1.99	96.0 [88.0–99.0]

1. ^aStandard deviation. ^bIncreased variability which amplitude range is >25 beats/min for 1 min or longer was defined as ZigZag pattern in the literature [23]. ^cShark-teeth-like increased uterine contractions.

Figure 1:

Minimal variability in the CTG trace of a patient with COVID-19 infection.

Figure 2:

Marked variability in the CTG trace of a patient with COVID-19 infection.

Figure 3:

Loss of accelerations with variable decelerations in the CTG trace of a patient with COVID-19 infection.

Figure 4:

Recurrent decelerations with minimal variability in the CTG trace of a patient with COVID-19 infection.

Figure 5:

Fetal tachycardia in the CTG trace of a patient with COVID-19 infection.

There was a statistically significant difference in terms of baseline FHR values between first and last CTG traces (p<0.05). Final baseline FHR values were significantly higher than the initial baseline FHR value.

There was no statistically significant relationship between COVID-19 severity and CTG category, CTG category change, variability, tachycardia, bradycardia, acceleration, deceleration, and uterine contractility (p>0.05) (Table 4).

Table 4:

Relationship between COVID severity and CTG features.

COVID-19 severity	Mild		Moderate/severe		p-Valuea
	n	%	n	%
CTG category					p=0.247
Category 1	123	71.7	27	77.1
Category 2	43	24.9	5	14.3
Category 3	7	4.0	3	8.6
CTG category changes					p=0.460
No	138	80.2	29	82.8
Category 1–2	10	5.8	2	5.7
Category 2–3	3	1.7	–	–
Category 1–3	7	4.1	1	2.9
Category 2– 1	12	7.0	1	2.9
Category 3–1	2	1.2	2	5.7
Variability					p=0.315
Absence of variability	8	4.6	–	–
Minimal variability	38	22.0	9	25.7
Normal variability	119	68.8	26	74.3
Marked variabilityb	8	4.6	–	–
Tachycardia/bradycardia					p=0.744
No	161	93.1	32	91.4
Tachycardia	10	5.8	2	5.7
Bradycardia	2	1.1	1	2.9
Acceleration					p=0.792
No	25	14.5	4	11.4
Yes	148	85.5	31	88.6
Deceleration					p=0.870
No	130	75.1	25	71.4
Early	3	1.7	1	2.9
Variable	23	13.3	5	14.3
Late	14	8.2	4	11.4
Prolonged	3	1.7	–	–
Uterine contractility					p=0.960
No	49	28.4	10	28.6
Yes	114	65.9	23	65.6
Tachysystole	7	4.0	1	2.9
Saw-toothc	3	1.7	1	2.9

1. ^aFisher-Exact/chi-square test. ^bIncreased variability which amplitude range is >25 beats/min for 1 min or longer was defined as ZigZag pattern in the literature [23]. ^cShark-teeth-like increased uterine contractions.

There was no statistically significant difference in terms of gestational week, first and last baseline FHR values, birth weight, Apgar 1st and 5th min between COVID-19 categories (p>0.05). A statistically significant difference was found in terms of the lowest oxygen saturation value according to COVID-19 categories (p<0.05). The lowest oxygen saturation values of the moderate and severe COVID-19 patients were statistically significantly lower than the mild ones (Table 5).

Table 5:

Relationship between COVID-19 severity and clinical features.

COVIDs Severity	Mild		Moderate/severe		p-Valuec
	Mean ± SDa	Median [IQRb]	Mean ± SDa	Median [IQRb]
Gestational week	36.40 ± 3.44	38.0 [4.8]	35.60 ± 3.30	37.0 [5.0]	p=0.079
Baseline FHR (first)	136.19 ± 11.91	140.0 [10.0]	134.57 ± 12.62	135.0 [15.0]	p=0.523
Baseline FHR (last)	138.37 ± 11.61	140.0 [15.0]	141.14 ± 11.76	140.0 [20.0]	p=0.158
Apgar 1st min	7.55 ± 0.81	8.0 [1.0]	7.17 ± 1.55	8.0 [1.0]	p=0.374
Apgar 5th min	9.09 ± 0.77	9.0 [1.0]	8.75 ± 1.03	9.0 [0.0]	p=0.106
Birth weight, g	3,101.03 ± 599.10	3,200.0 [585.0]	2,928.80 ± 608.59	3,020.0 [650.0]	p=0.135
Min O2 saturation	96.78 ± 1.17	97.0 [2.0]	93.43 ± 2.75	94.0 [4.3]	p=0.000d

1. ^aStandard deviation. ^bInterquartile range. ^cMann-Whitney U test. ^dStatistically significant (<0.05).

As a result of the backward logistic regression analysis, utilizing all available parameters and the COVID-19 severity, the optimal model was created and presented in Table 6.

Table 6:

Logistic regression model based on COVID-19 categories.

Variable	p-Value	ORa	95% Confidence interval (OR)
			Low	High
Gestational week	0.059	0.898	0.803	1.004
CTG categorya	0.125
Category 2	0.107	0.375	0.114	1.235
Category 3	0.253	2.363	0.541	10.320
Baseline FHR (first)	0.034b	0.955	0.915	0.996
Baseline FHR (last)	0.023b	1.053	1.007	1.101

1. ^aOdds ratio. ^bStatistically significant (<0.05).

It was observed that the COVID-19 categories affected the final baseline FHR value. The baseline value increases by one beats/min (bpm), when the risk of being COVID-19 moderate or severe increases by 5.3% (OR=1.053).

Discussion

To the best of our knowledge, this is the first study with a significantly high number of patients which analyzed the CTG changes in maternal COVID-19 infection using three-tier FHR system [22] and documented all CTG trace parameters according to the COVID-19 severity.

Numerous factors or circumstances have influence on the characteristics of the FHR pattern, including the time of the day, position and activity of the mother, movement of the fetus [24], [25], [26], [27]. In the current study, we chose the optimal timing, activity and position to maximize oxygenation of the fetus: postprandial, in the morning, at rest, and semi-sitting position.

Maternal COVID-19 infection can cause changes that can be observed in CTG due to reasons such as maternal hypoxia, cytokine storm, maternal pyrexia, uterine irritability, and placental intervillous thrombosis secondary to maternal hypercoagulable state, which may result with diminished transfer of oxygen through the placenta [10], [11], [12], [13], [14], [15, 28].

Maternal inflammatory state and pyrexia is likely responsible from main CTG changes in COVID-19 infected mothers such as increased baseline FHR >10% and fetal tachycardia (FHR >180 bpm) (Figure 5). Inline with the results of the only study performed with 12 patients [29], in our study, all fetuses showed an increased baseline FHR compared to the initial recording. Traces with fetal tachycardia were mostly mild cases, but two severe COVID-19 cases had FHR >200 bpm which was most likely secondary to the maternal cytokine storm due to excessive inflammatory mediators and fetal sympathetic response.

Maternal hypercoagulable state, maternal hypoxia and ARDS result with placental intervillous thrombosis and reduction in placental circulation which can cause acute fetal bradycardia in CTG trace [30, 31]. After acute fetal bradycardia was seen in a mother with severe COVID-19, the fetus was delivered with an emergency C/S operation in which the newborn had 1st min APGAR 7. The oxygen saturation was 88 without the nasal oxygen support and she went under invasive ventilation after delivery. However, at the time of CTG recording, the patient received continuous nasal oxygen support and the oxygen saturation was above 92.

Although most of the patients had normal variability, there were also cases with minimal (48 patients) and increased (nine patients) variability in our study because the population size was high enough. Maternal hypoxia and cytokine storm associated inflammatory cytokines cross the placenta and the fetal blood-brain barrier, and cause depression of fetal central nervous system [29]. This is observed as minimal variability (<5 bpm) in CTG (Figure 1). Majority of our population had normal variability (159 patients). Inflammatory cytokines and maternal pyrexia can cause fetal autonomic instability. This is observed as increased variability (>25 bpm) which results with ZigZag pattern in CTG trace. In our study, on the contrary to Gracia-Perez-Bonfils A et al. [29], ZigZag pattern (Figure 2) was only seen in patients who were in mild category. No increased variability was observed in moderate and severe categories.

Maternal hypoxia and inflammatory cytokines can cause depression of fetal central nervous system which results with reduction of fetal movements and seen in CTG as loss of accelerations [30, 31]. In our study, not associated with COVID-19 severity, 30 patients showed loss of accelerations (Figure 3). Despite the majority of the patients had no decelerations, all types of decelerations especially variable decelerations were seen in CTG trace (Figures 3 and 4), and again exhibiting no association with the COVID-19 category. Although our study design prevents us from speculating about the cause of these decelerations, these results support our theory. Maternal hypoxia or hypotension and maternal hypercoagulable state leading to placental thrombosis and infarction which cause increased placental metabolism and uteroplacental insufficiency may be the reason of decelerations seen in CTG trace [7, 14, 15].

Rarely, sinusoidal pattern can be observed in CTG trace as a result of chronic fetal anemia and acidosis due to both maternal and fetal destruction of red cells. Fetal autonomic instability caused by acute feto-maternal hemorrhage and fetal hypovolemia can also cause the pattern. Similar to the previous study [29], sinusoidal pattern was not seen in CTG traces.

Myometrial irritability due to maternal inflammatory state and maternal pyrexia is associated with increased uterine activity [16]. As expected, in our study increased uterine activity was observed in majority of the patients.

The main strength of our study is the high number of patients analyzed in the study population. While the only previous study analyzing the CTG features of COVID-19 pregnancies recruited 12 patients, we worked with 224 patients. Another major strength is that this is the first characterization of detailed CTG findings in women with acute COVID-19 infection.

The main limitations of our study are that study design does not allow determination of causality, and the number of moderate and severe COVID-19 cases is low. However, although the numbers were small in moderate and severe cases comparatively, since the population size is high, the small numbers are considered sufficient to present early results about COVID-19 during pregnancy. Other limitations are the lack of correlation between the CTG trace features and the newborn umbilical cord pH, and the lack of knowledge about the possible effect of the cocktail of medications on the CTG.