Characterization of Cervical Remodeling During Pregnancy and the Postpartum Period: A Case Series Analysis

Introduction

The uterine cervix plays an important role in maintaining pregnancy to term. In a normal, full-term pregnancy, when the time is right, hormonal cues instruct the cervix to soften (ie, biomechanical changes in early pregnancy that are marked by a gradual reduction in tissue stiffness while maintaining tensile strength), ripen (ie, biomechanical and structural changes late in pregnancy that lead to the loss of tensile strength and increase tissue hydration), and dilate for delivery [1,2]. Alterations in the timing of cervical remodeling can result in asymptomatic cervical changes that go undetected and lead to premature cervical dilation and spontaneous preterm birth (sPTB) [3]. After delivery, cervical repair is the last phase in the cervical remodeling process and is responsible for restoring tissue integrity and function [1,2]. Despite its important role in pregnancy and future reproductive health, the process of cervical remodeling in pregnancy and the postpartum period is not well understood, and currently, there is no reliable method to detect abnormal or impaired cervical remodeling.

Characterization of the cervical remodeling process in humans has been challenging due to the lack of noninvasive options, which has resulted in limited longitudinal studies evaluating the process from pregnancy to the postpartum period. Current techniques to evaluate cervical remodeling, specifically collagen content and organization, primarily rely on cervical biopsy. Cervical biopsy is painful and carries a risk of bleeding, infection, and physiological distress [4,5]. Also, cervical biopsy is rarely performed for research during pregnancy, due to the lack of clinical necessity [6]. Moreover, standard biopsy samples are typically small and may not capture changes in the deep stroma of the cervix, limiting their utility. Quantitative ultrasound (QUS) is a repeatable, noninvasive ultrasound technology that can detect subtle, subclinical changes in tissue properties [7,8]. QUS analyzes how the ultrasound waves interact with tissue and is sensitive to changes in cervical collagen organization, hydration, and stiffness [9,10]; thus, it has the potential to noninvasively characterize the cervical remodeling process longitudinally from pregnancy through the postpartum period. Ultrasound is available in almost every obstetric practice, and QUS could be an added feature to current clinical ultrasound systems. Unlike conventional ultrasound, which assesses macroscopic features, including cervical length, fetal anatomy, placental position, and amniotic fluid volume [11], novel QUS enables quantification of the 5 following biomarkers that offer objective information about the tissue microstructure: (1) attenuation coefficient measures the amplitude loss of the ultrasound wave as it passes through tissue and reflects tissue composition and hydration; (2) backscatter coefficient quantifies the amount of energy scattered back to the transducer and is sensitive to the size and spacing of microstructural components (eg, collagen fibers and collagen fiber alignment); Lizzi-Feleppa (3) slope and (4) intercept, which are based on how ultrasound backscatter signals change with frequency, can be used to infer information about tissue microstructure and collagen remodeling (eg, changes in how collagen fibers are organized and spaced) by capturing how waves scatter off of tissue structures; and (5) shear wave speed, which is derived from shear wave elastography and is an indirect measure of tissue stiffness (where higher speeds reflect stiffer tissue) [7,9,10,12–16]. While their use has been limited, novel QUS biomarkers offer a promising noninvasive method to monitor tissue level changes from pregnancy to postpartum recovery.

In this case series analysis, we used QUS to assess cervical remodeling from 2 time points in mid-pregnancy (approximately 20 and 24 weeks of gestation) and 5 time points in the postpartum period (approximately 6 weeks, and 3, 6, 9, and 12 months postpartum) to characterize the cervical remodeling process in 8 participants.

Case Reports

STUDY PROCEDURES:

Participants were provided information about ultrasound examinations, including that they are a minimal-risk procedure, and that QUS data acquisition poses no additional risk. The thermal index and mechanical index were maintained below the US Food and Drug Administration recommended safety thresholds, and total scanning time was limited to 5 min. Participants were counseled on the minimal risks of undergoing transvaginal ultrasound examinations, including the possibility of an allergic reaction to latex or the coupling gel and/or experiencing discomfort during the transvaginal ultrasound examination. After providing consent, participants were asked to empty their bladder. Participants were provided a cotton drape sheet, instructed to remove their clothing from the waist down, and given privacy to prepare for the examination. Once the participant was ready, they were asked to lay in a supine position with a pillow supporting their head and neck. The examination was performed on an examination table with stirrups and pelvic tilt. The transvaginal scans were performed using a Siemens Acuson S2000 (Siemens Healthineers, Munich, Germany) diagnostic ultrasound system with one of four MC 9-4 transvaginal ultrasonic transducers (measured frequency range, 3.7–6.8 MHz; center frequency, 5.25 MHz). Each participant underwent a total of 11 S2000 acquisitions (10 cervix and 1 reference phantom acquisition for calibration purposes) at each of the 7 time points. The detailed protocol for the transvaginal scans and QUS processing are described in detail in our previous studies [17–19].

We selected 5 QUS biomarkers to assess changes in the collagen microstructure of the cervix. These included attenuation coefficient (dB/cm-MHz); backscatter coefficient (dB); and Lizzi-Feleppa biomarkers slope and intercept (dB) [10,12,13]. The biomarker shear wave speed (m/s) was also collected at the time of scanning from a sonographer-defined region of interest (selected at the time of scanning), and no offline processing was required [14,16]. Together, these biomarkers can provide insights into the cervical tissue’s biological composition, microstructure, and mechanical properties. These biomarkers were selected because findings from an animal study have demonstrated that QUS biomarkers are related to collagen concentration, collagen disorganization, water content, and gestational age changes in the cervix [9]. Human studies have validated that these QUS biomarkers are sensitive to collagen changes in the cervix [8,20]. While there is availability of other QUS biomarkers, including strain elastography to assess cervical changes in non-pregnancy and pregnancy and in response to misoprostol administration, shear wave speed was selected because it provides a quantitative measure of tissue stiffness with minimal operator dependence, making it well suited for the longitudinal evaluation of cervical remodeling [21–24].

CASE DATA AND BIOMARKER TRENDS:

Individual participant characteristics are summarized in Table 1. All data in this results section are given as mean (SD). The mean (SD) age at enrollment was 31.6 (5.8) years. One participant was primigravida, and 7 were multigravida. Six pregnancies resulted in full-term births, with a mean gestational age of 39w2d, and 2 participants delivered preterm, with a mean gestational age of 34w4d. The mean infant birth weight was 3043.5 (463) g. Three participants did not breastfeed, 1 breastfed for less than 3 months, and 4 were breastfeeding at 12 months. At 12 months postpartum, 7 of 8 participants had resumed menstruation.

Of the 5 QUS biomarkers analyzed, attenuation coefficient, backscatter coefficient, and slope showed similar trends across the 7 visits. For attenuation coefficient, the mean (SD) values decreased from 1.03 (0.25) dB/cm-MHz at V1 to 0.85 (0.13) at V2 and 0.80 (0.22) at V3, then increased to 1.14 (0.22) at V4, followed by a decrease to 0.91 (0.35) at V5 and subsequent increases to 1.01 (0.19) at V6 and 1.09 (0.26) at V7. Backscatter coefficient followed a comparable pattern, declining from −18.74 (3.02) dB at V1 to −20.29 (2.56) at V2 and −24.92 (2.45) at V3, rising to −21.35 (2.43) at V4, followed by a small decrease to −23.79 (3.26) at V5 and increased to −22.71 (2.75) at V6 and −22.45 (2.83) at V7. Slope decreased from 2.84 (0.72) dB/MHz at V1 to 2.40 (0.29) at V2 and 2.22 (0.59) at V3, rose to 2.79 (0.48) at V4, decreased to 2.21 (0.88) at V5, and increased again to 2.60 (0.79) and 2.58 (0.56) at V6 and V7, respectively.

Shear wave speed values were lower during pregnancy, with a mean of 1.6 (0.3) m/s. After delivery, these values increased to a mean of 3.2 (0.2) m/s. Intercept values were higher at both pregnancy time points, with a mean of −33.3 (0.52) dB, and decreased postpartum to a mean of −36.1 (0.45) dB.

Figure 1 illustrates the QUS biomarker values as boxplots, displaying the distribution from the 10 repeated scans for each participant across the 7 visits.

Discussion

QUS is an ultrasonic method that estimates tissue properties by quantifying the interaction between ultrasound waves and biological tissues. QUS biomarkers reflect the changes in ultrasonic attenuation and scattering based on tissue composition and organization and may provide insights into the progression of cervical tissue remodeling. In this case series, our findings suggest that from approximately 20 weeks’ gestation to 6 weeks postpartum there is a gradual decrease in cervical attenuation coefficient, backscatter coefficient, and slope. Findings from animal and human studies [8–10] suggest that the decrease in these biomarkers indicates that there is increase in water concentration and collagen disorganization and a decrease in collagen concentration, consistent with cervical softening during pregnancy. We hypothesize that the decrease observed at 6 weeks postpartum may be related to the inflammatory and early proliferative phases of wound healing, consistent with studies that have demonstrated that cytokines and matrix metalloproteinases are increased postpartum, potentially facilitating collagen degradation and reorganization during a period when newly synthesized collagen is still immature and disorganized [25–28]. At 3 months postpartum, there is an increase in these QUS biomarkers. Animal studies have demonstrated upregulation of genes involved in collagen synthesis in the postpartum period, we hypothesize that these changes may reflect ongoing cervical repair in humans characterized by collagen fibril organization and maturation [25,28,29]. At 6 months postpartum, attenuation coefficient, backscatter coefficient, and slope decrease; we hypothesize this corresponds to the maturation phase of wound healing, in which excess or immature collagen is broken down and removed [25]. By 9 months postpartum, there was a small increase from the 6-month biomarker values, which remained consistent with the 12-month biomarker values. We postulate this reflects a leveling off or stabilization of cervical remodeling, consistent with findings from a study [30] that found collagen concentration significantly increases until 9 months postpartum and stabilizes by 12 months postpartum.

Shear wave speed values decreased as pregnancy progressed (approximately 20 and 24 weeks of gestation), reflecting the softening phase of cervical remodeling, during which there is an increase in water content, and collagen becomes more flexible and less rigid, resulting in lower mechanical strength in preparation for delivery [14,31]. During this process, increased spacing between the collagen fibrils has been observed, and the cervix undergoes remarkable changes in its biostructure, biochemistry, and biomechanics as it transitions from an unripe to a ripe state [8,31,32]. At 6 weeks postpartum, there is a sharp increase in shear wave speed that is sustained throughout the postpartum period, suggesting that the tissue becomes stiffer and more rigid. It is known that there is a rapid return to normal cervical tissue architecture in the postpartum period [33], strengthening the hypothesis that collagen in the cervix becomes more tightly packed, denser, and well-aligned (organized), with restored mechanical strength.

Intercept values were higher during both pregnancy time points, at approximately 20 and 24 weeks of gestation. Higher intercept values suggest that the size and concentration of scatterers are increased [18,34]. We hypothesize that higher intercept values are influenced by changes in the composition of the extracellular matrix. During pregnancy, proteoglycans and glycosaminoglycans increase as pregnancy progresses [35,36]. The glycosaminoglycans may play a role in altering the shape of collagen fibers under strain and contribute to the mechanical changes observed during pregnancy [31,37]. After birth, intercept values decreased at the 6-week postpartum scan and remained stable for the remaining postpartum scans. We hypothesize that alterations in the extracellular matrix composition and structure, including a decrease in glycosaminoglycans and other matrix components could lead to the observed decline in intercept values.

Conclusions

This case series presents QUS measurements observed in 8 participants from mid-pregnancy to 1 year postpartum. We hypothesize on potential structural and functional tissue level changes occurring in the cervix during this period. Continued research in cervical remodeling using QUS is needed and has the potential to inform clinical practice based upon cervical tissue changes in pregnancy and the postpartum period. Advancements in this technology could allow for earlier detection of individuals at risk for preterm birth, even in the absence of known risk factors or clinical symptoms. It may also help identify those with incomplete cervical remodeling that could affect future pregnancies, ultimately supporting the development of more timely and targeted interventions. Because QUS technology can be programmed into existing ultrasound systems that are routinely used in pregnancy, this technology could be accessible to any facility with an ultrasound machine and would reduce the time from research to bedside use. Overall, these 8 cases offer valuable insights into the process of cervical remodeling throughout pregnancy and the postpartum period and can serve as a basis to assess cervical remodeling in larger-scale clinical studies.