Role of mitochondrial DNA level in epidural-related maternal fever: a single-centre, observational, pilot study

Epidural analgesia is commonly used to relieve labour pains, and labour in 20% of women is associated with increased body temperature [1]. The underlying epidural-related maternal fever (ERMF) mechanisms remains unclear although the causal relationship between maternal fever and epidural analgesia is well-known [2, 3].

A current hypothesis is that ERMF is a non-infectious inflammatory reaction that is putatively triggered by damage-associated molecular patterns (DAMPs), which can activate the inflammasome [4]. Interestingly, several in vitro studies demonstrated that local anaesthetics have the potential to trigger immunomodulation and cell injury due to mitochondrial respiratory chain complex inhibitions [5, 6]. Consequently, mitochondrial damage induces the release of large amounts of reactive oxygen species, which promote acute systemic inflammation via pro-inflammatory mediator production [e.g. cytokines, prostaglandin E2, oxidated mitochondrial deoxyribonucleic acid (mtDNA)] [7, 8]. Additionally, mitochondrial stress induces the escape of mtDNA into the extracellular space, which might not be necessarily a passive release, but a specific and regulated process via several possible routes [9].

Studies on various disease states demonstrated that circulating cell-free mtDNA alterations can be measured in peripheral blood, which reflects the degree of mitochondrial injury. Although, for preeclampsia mtDNA dysregulations have been described by different study teams [10, 11], no in vivo study has yet investigated the correlation between cell-free mtDNA alterations and ERMF. Different experimental approaches revealed that local anaesthetics, such as bupivacaine and (to a lesser extent) ropivacaine, have an impact on mitochondrial function, induce ROS generation and are dose-dependently pro-apoptotic [5].

Therefore, this pilot study aimed to investigate alterations of plasma circulating cell-free mtDNA levels in women in labour with ERMF in comparison to women in labour with epidural analgesia an no fever and women without epidural analgesia.

This pilot study was conceived in preparation for further clinical studies, and the results demonstrate that women with ERMF during labour exhibit altered mtDNA dynamics. Furthermore, plasma IL-6 levels significantly increased during labourin women with an epidural catheter.

Confirming previous studies, 20% of women with epidural analgesia developed a temperature increase of > 37.5 °C during labour in this pilot trial [4, 18]. The intrapartum temperature elevation started within 4–8 h after epidural placement. Our study did not reveal an association with a higher body mass index or labour duration as previously reported [18]. However, PROM was associated with ERMF as several studies previously indicated [19]. IL-6 level elevation in plasma from women with epidural analgesia was also previously found in a study by Riley et al. [20].

The course of physiological and molecular events that lead to ERMF remained controversial; however, recent data indicate that maternal pyrexia in the absence of infection involves sterile inflammation processes and/or altered thermoregulation [1, 20]. Several clinical and experimental studies have provided evidence of possible inflammatory mechanisms, including the direct effects on inflammatory cytokine release [17], anaesthetic-induced mitochondrial disturbances, reactive oxygen specie generations and pro-apoptotic developments [5].

The release of mtDNA from mitochondria into the cytoplasm and further into the extracellular space is proposed to occur via different mechanisms, which is not understood in every detail [21]. Cell-free mtDNA levels were increased towards term during healthy pregnancy and returned to initial lower levels 6–8 several weeks after giving birth [22]. However, mtDNA dysregulations have been described during pregnancy in women with preeclampsia resulting in different profiles reported by different study teams, as well as transitional suppressed mtDNA levels [23]. Different experimental approaches revealed that local anaesthetics, such as bupivacaine and (to a lesser extent) ropivacaine, have an impact on mitochondrial function, induce ROS generation and are dose-dependently pro-apoptotic [5]. These findings have prompted us to investigate if secreted mtDNA could be a contributor to inflammatory processes that lead to ERMF. Interestingly, this pilot trial did not reveal increased plasma mtDNA levels in women with ERMF. However, patients with ERMF showed mtDNA dysregulation compared with women without ERMF, who had a stable situation of circulating cell-free mtDNA, during labour. The ERMF group exhibited a decreased circulating cell-free mtDNA during labour, which was statistically significant, and a (non-significant) tendency to higher cell-free nDNA levels. The emerging pattern is similar to what was observed by Cushen et al. in women with preeclampsia, although to a lesser extent, which might probably be due to the lower number of women in our ERMF group [23]. Our study revealed a higher amount of oxidated cell-free mtDNA in the ERMF group at T0; however, it did not reach statistical significance due to the large variation of individual values.

The results of this pilot trial cannot support the theory that ERMF may be explicitly traced back to a local anaesthetic-induced mitochondrial damage and concomitantly altered mtDNA release dynamic; however, the findings demonstrate the importance of future investigations in the field of circulating cell-free mtDNA and sterile inflammation processes in the context obstetric anaesthesiology.

This pilot study was performed in preparation of further clinical studies. One finding in this study was that the transfer from bench to bedside needs more evidence. Due to challenging recruitment of patients with ERMF following clinical studies would benefit from multicenter studies. Furthermore, due to the fact that the timepoints for the determination of mtDNA alterations even aspects during pregnancy should be considered and moreover, non pregnant patients with an epidural analgesia and even patients with a caesaran section should even be observed as control groups.

A major limitation is that due to missing data in this research field no formal sample size calculation could be possible. The defined sample size for for the required women with ERMF in this pilot study were based on a similar work. However, the present data of this pilot study can serve as a basis for designing future clinical trials in this area. Furthermore, limitations include the small sample size besides diversity and ethnicity. Moreover, blood samples were only taken after admission to the delivery room and immediately after birth. Circulating cell-free mtDNA levels during pregnancy and labour progress would be of major interest. In addition a limitation of the method is that only non-hemolytic samples can be analysed. Due to this fact mtDNA levels could not be measured in placental blood.

Importantly, epidural analgesia for women in labour was given as a ropivacaine/fentanyl combination in this pilot study, whereby fentanyl could be a confounding factor concerning the study outcome. Previous studies revealed that fentanyl has an impact on the mitochondria [24] and can decrease the mitochondrial copy number in blood although only in male and not in female mice [25]. However, the significantly decreased cell-free mtDNA only in the ERMF group speaks against this obstacle, because the control group of ‘EA no fever’ that received the anaesthetic combination did not show the effect.

The use of Ct-values to compare groups might appear to the reader as if authors have not further processed their raw data in order to perform calculations of "fold change" relative to the control group. However, this decision was made for the following reason: circulating cell-free nuclear DNA and mtDNA are most probably released from cells independently by different mechanisms. Cell-free nuclear DNA is released from cells undergoing cell death (apoptosis), while mtDNA can be released in other context (cell stress or other triggers). An indicator, that this might also be the case in this study, is shown by the fact, that in the ERMF group the mtDNA shows a downward trend between the different time points, while cf nDNA shows an upward trend. Therefore normalizing cf mtDNA levels to cf nDNA would induce a bias. As we have used the same sample size in all cases and have processed all samples according to the same protocol, we decided to use the raw Ct values to compare the data. This procedure makes the differential effect on cf nDNA and cf mtDNA more visible.

Up to 20% of women in labour with epidural analgesia might be affected by maternal fever; however, the underlying cause of ERMF is not fully understood. The present study demonstrates cell-free mtDNA dysregulation in women with maternal fever. The results support the theory of sterile inflammation in ERMF and support the urgent need for future investigations, in vivo and in vitro, to elucidate the origin of the ERMF phenomenon.