Trajectory of Postpartum Cervical Remodeling in Women Delivering Full-term and Spontaneous Preterm: Sensitivity to Quantitative Ultrasound Biomarkers

Abstract

Objective:

Women with a history of spontaneous preterm birth (sPTB) face an increased risk of recurrence. Yet, the factors contributing to the increased risk are unknown, hampering the development of targeted interventions. Noninvasive Quantitative Ultrasound (QUS) has been validated in the characterization of cervical tissue and has the potential to provide information about postpartum cervical remodeling. The objective of this study was to determine the postpartum cervical remodeling trajectories of women over 12 months post-delivery and to determine whether there were differences between women who delivered full-term and spontaneous preterm that were sensitive to QUS biomarkers.

Methods:

Data were collected prospectively from 55 women: 41 who delivered full-term and 14 who delivered spontaneously preterm at 6 weeks, 3, 6, 9, and 12 months (±2 weeks) postpartum. Data from QUS biomarkers: Attenuation Coefficient; Backscatter Coefficient; Shear Wave Speed; and Lizzi-Feleppa Slope, Intercept, and Midband were analyzed from the acquired radiofrequency data using a Siemens S2000 ultrasound system with a transvaginal MC 9–4 MHz probe. The biomarkers were analyzed using descriptive statistics and linear mixed-effects models.

Results:

QUS biomarkers, Backscatter Coefficient, and Lizzi-Feleppa Intercept showed significant differences during the year after delivery between women who had a full-term birth and sPTB (p<0.05), suggesting that there are differences in the cervical remodeling trajectories between the two groups. All QUS biomarkers demonstrated significant variations between the full-term birth and sPTB groups over time (p<0.05), indicating ongoing cervical remodeling for both groups during the 12-month postpartum period.

Conclusion:

QUS biomarkers identified cervical microstructure differences and trajectories in the year after delivery between women who delivered full-term and spontaneous preterm.

INTRODUCTION

In the United States, 10.4% of infants were born preterm in 2022.¹ Infants born preterm have higher rates of lifelong neurodevelopmental, functional, and behavioral complications,^2,3 at a cost to society of more than $25.2 billion annually.⁴ Preterm birth (PTB) is defined as a live birth occurring prior to 37 weeks of gestation.⁵ Preterm birth is further categorized as medically indicated (due to maternal or fetal factors) or spontaneous (idiopathic or due to preterm premature rupture of membranes (PPROM)).^6,7 Spontaneous preterm birth (sPTB) accounts for 70% of all PTBs⁷ and is the most significant risk factor for having another sPTB.^8,9 Having two or more sPTBs is defined as recurrent sPTB.¹⁰ According to Phillips et al., women who have delivered spontaneously preterm have a 30% absolute risk of recurrence in future pregnancies (95% confidence interval 27% to 34%)¹¹ and the reason for this increased risk is unknown. Therefore, it has not been identified which women are at risk for recurrence, which hinders the development of targeted interventions.

Preterm birth (PTB) is a complex syndrome with multiple potential causes, including infection (clinical or occult), hormone imbalance, alterations in immune function, vascular disorders, breakdown of maternal-fetal tolerance, and cervical disorders.^12,13 Regardless of the underlying etiology or timing of labor onset (full-term and preterm), the final common pathway involving cervical remodeling and then dilation must occur because contractions alone are insufficient to affect cervical change and delivery.¹⁴ The uterus comprises two fundamentally distinct regions: the corpus (body) and the cervix. The corpus consists primarily of smooth muscle, while the cervix consists primarily of collagen (approximately 70% type I and 30% type III).^15–17 Findings from animal and human studies during pregnancy showed that collagen organization in the cervix contributes to its mechanical strength and therefore, plays a role in maintaining the pregnancy.^16,18–21 As pregnancy progresses to term, the presence of matrix metalloproteinases and inflammatory cytokines increases, water content increases, and there is a decrease in hydroxyproline, which is a major component of collagen, the structural protein that provides stability.^16,22,23 Consequently, there is a loss of cross-linked collagen, resulting in disorganized collagen and a decrease in collagen concentration in the cervix.^{16,21,24–26} Thus, the cervix shortens and softens, and allows passage of the fetus through a less rigid cervix.^16,27 In a preterm pregnancy, it is known that the timing and triggers that allow the cervix to remodel are different than a full-term pregnancy.^28–31 Alterations in the normal cervical remodeling process can result in cervical microstructural changes that lead to funneling (opening of the internal portion of the cervix), premature cervical dilation, PPROM, and preterm delivery.^16,25

Findings from animal studies showed that postpartum cervical remodeling begins immediately after delivery and involves several carefully orchestrated processes, including a robust inflammatory response.^32,33 The process of postpartum cervical remodeling is poorly understood; but, it is postulated that an unregulated response could lead to cervical damage and altered wound healing leading to an increased risk of recurrent sPTB.^33,34 In one human study³⁵ evaluating collagen concentration after birth, investigators reported that after a full-term delivery, there was a gradual and significant increase in collagen concentration for up to nine months postpartum. In another human study,³⁶ investigators determined that women who had suffered from cervical insufficiency (premature, painless dilation of the cervix) during pregnancy that resulted in late spontaneous abortion or PTB had lower collagen concentrations than women without cervical insufficiency. To our knowledge, no studies have evaluated cervical remodeling in postpartum women in the first year after a preterm delivery. Additionally, the studies that have evaluated postpartum cervical remodeling have used cervical biopsy samples.^35–37 Cervical biopsies are painful,³⁸ invasive, and have not been conducted in routine clinical care to evaluate collagen and cervical healing in the United States.

Due to the lack of noninvasive technologies and focus on preterm deliveries during pregnancy, only a few studies have evaluated cervical remodeling in the postpartum period.^35,39 For the first time, we can measure the microstructural properties of the cervix noninvasively with Quantitative Ultrasound. Quantitative Ultrasound is a noninvasive ultrasound technology to quantify tissue microstructure and function.⁴⁰ Quantitative Ultrasound has been validated in animal and human studies^{19,26,30,41,42} to characterize cervical tissue properties (cervical collagen remodeling) during pregnancy. Given that there is a limited understanding of the factors that lead to the increased risk of recurrent sPTB, it is valuable to determine if the microstructure changes that occur in sPTB persist in the postpartum recovery period and create long-standing changes in the cervical tissue that increase the risk of recurrent sPTB.

The objective of this study was to determine the trajectory of cervical remodeling at 6 weeks, 3, 6, 9, and 12 months postpartum in women who delivered full-term and spontaneous preterm, and to determine whether there were differences in the cervical remodeling trajectories that were sensitive to Quantitative Ultrasound biomarkers. We hypothesized that women who delivered spontaneous preterm would have distinct cervical remodeling trajectories, characterized by differences in collagen content and organization when compared to women who delivered full-term, and that these differences would be sensitive to Quantitative Ultrasound biomarkers.

MATERIALS AND METHODS

In this prospective cohort study, participants were recruited from the Center for Women’s Health at the Outpatient Care Center at the University of Illinois Chicago. All participants self-identified as female at time of birth and enrollment. Women were enrolled in the study between October 2019 and May 2022 during a prenatal visit or at their 6-week postpartum visit. Study approval was obtained from the Institutional Review Board at the University of Illinois Chicago (UIC). Written informed consent was obtained prior to the first postpartum study scan. This study was a sub-study of a large, ongoing preterm birth risk prediction study.^42,43 Women were not exclusively recruited from the parent study; however, the inclusion and exclusion criteria were predefined by the parent study. Women were eligible to participate in the study if they were ≥15 years, had a singleton pregnancy, and could speak, read, and write English. Women were excluded if they had any of the following conditions during pregnancy: chronic medical conditions (chronic hypertension, HIV, preeclampsia, gestational diabetes, Type I diabetes, or autoimmune disorder), major fetal anomaly, multiple gestation, had received steroids and/or anti-inflammatory treatment for a preexisting chronic condition or had a routine cesarean section.

Recruitment and Retention

All research activity was halted from March 2020 until February 2021 due to Coronavirus disease 2019 (COVID-19). When COVID-19 restrictions were lessened, women were still hesitant to join or maintain scheduled follow-up postpartum scans. Reasons cited included not wanting to increase exposure to COVID-19 by coming to the hospital, a lack of childcare, and limited travel and transportation options. Additionally, some patient scans were not conducted within the scheduled timeframe because of delays due to repeated rescheduling, missed appointments, relocation, or COVID-19 infection.

Quantitative Ultrasound Selected Features

Quantitative Ultrasound is a noninvasive ultrasound technology used to quantify tissue microstructure.^19,26,30,41 The parameters derived from Quantitative Ultrasound serve as biomarkers, as defined by the Quantitative Imaging Biomarkers Alliance (QIBA), due to their ability to quantitatively assess tissue characteristics on a ratio or interval scale. These provide precise markers for normal and abnormal biological processes or responses to therapeutic interventions.⁴⁴ The following Quantitative Ultrasound biomarkers were used to evaluate the cervical microstructure changes in the postpartum cervix in women who delivered full-term and sPTB: Attenuation Coefficient in units of dB/cm-MHz, the quantified reduction in the strength of an ultrasound wave as it propagates through tissue; Backscatter Coefficient in units of 1/sr-cm that were log-transformed [i.e., 10 log₁₀(BSC)] into the dB scale, and denoted as Backscatter Coefficient (dB), a quantitative measure of the echogenicity; Shear Wave Speed in units of m/s, a measure of tissue elastic properties; and the three Lizzi-Feleppa biomarkers: Slope (in units of dB/MHz), Intercept (in units of dB), and Midband (in units of dB) which are calculated using linear regression by analyzing the relationship between Backscatter Coefficient (dB) and frequency.^40,45–48

The Quantitative Ultrasound biomarkers describe distinct aspects of tissue properties: Attenuation Coefficient is primarily determined by the tissue biological composition; Backscatter Coefficient is primarily determined by the tissue microstructure;^49,50 Shear Wave Speed is a measure of tissue stiffness; Lizzi-Feleppa Slope provides information about the scatterer size distribution and density of tissue, Lizzi-Feleppa Intercept is related to the size, concentration, and the acoustic resistance of tissue scatterer and is a computational parameter derived from the backscattered ultrasound signals; and Lizzi-Feleppa Midband is the linear regression’s Midband value that is directly related to Backscatter Coefficient (dB).⁵¹ These Quantitative Ultrasound biomarkers were selected because findings from research in the rat cervix showed that Quantitative Ultrasound is related to collagen concentration, collagen disorganization, water content, and gestational age changes of the cervix.¹⁹ Findings from human studies showed that Quantitative Ultrasound identified women at risk for PTB.^30,42,43 Our group developed rigorous methods that have been validated in human and rat cervical tissue that demonstrated Quantitative Ultrasound is sensitive to collagen changes in the cervix.^19,26,30,41 We also established sonographer and ultrasound system reliability for these measures by assessing the repeatability and reproducibility of transvaginal Qualitative Ultrasound using a phantom-based approach, which showed the methods used across our studies were repeatable, and reproducible among sonographers, transducer probes, and probe covers.⁵²

Data Acquisition

Transvaginal scans were performed using the Siemens S2000 (Siemens Healthineers, Munich, Germany) diagnostic ultrasound system. One of four Siemens MC9–4 transvaginal ultrasonic transducers (measured frequency range: 3.7 – 6.8 MHz; center frequency 5.25 MHz) was used for each scan. The scans were performed by a member of the clinical team trained in the research protocol (MVD, 1 year of experience, RN; EDK, >3 years of experience, RDMS, RVT; SG, >3 years of experience, RDMS, RVT; TAP, >10 years of experience, CCRP; BLM, >31 years of experience, RDMS).

Transvaginal Quantitative Ultrasound scans were offered to each study participant at approximately 6 weeks, 3, 6, 9, and 12-months postpartum. A total of 11 S2000 acquisitions (10 cervix acquisitions and 1 phantom acquisition) were acquired for each participant at each of the 5 postpartum timepoints. For each visit, women were asked to empty their bladder before the scan. The transvaginal probe cover was filled with ultrasound gel and placed over the transducer; all air bubbles were removed by smoothing the cover over the transducer. The transvaginal probe was inserted vaginally without placing pressure on the cervix. The S2000’s ultrasound system settings were adjusted for cervical B-mode image optimization with a depth of approximately 6 cm and the focus immediately to the posterior of the lip of the cervix; all settings remained the same for the 11 acquisitions at each visit.⁵² A sagittal image of the cervix where the internal and external os were present and the entire uterus was visualized were obtained (Figure 1). Each acquisition recorded the B-mode (jpeg) image and raw radiofrequency data. Each acquisition was taken 8–15 seconds apart to allow the system to cool down in between each image capture. Once the 10 Quantitative Ultrasound cervix acquisitions were captured, the last (eleventh) acquisition was a calibrated reference phantom (CIRS, Inc., Norfolk, VA) with known Quantitative Ultrasound properties. The reference phantom was used as a Quantitative Ultrasound calibration factor and was specially made for use with the vaginal probe.⁵² The reference phantom was scanned immediately after the cervix was scanned. All scans were transferred to a personal computer (PC) for offline Quantitative Ultrasound processing.

Figure 1.

B-mode image of the cervix at 3 months postpartum.

Note: Pink line traces the cervical field of interest for QUS processing. Colored pixels are averaged to determine the mean SWS.

Quantitative Ultrasound Data Analysis

The acquired ultrasound data were processed offline on a PC using custom MATLAB (Mathworks, Natick, MA, USA) programs.^53,54 Sonographers (MVD, TAP, and BLM) drew a free-hand field of interest on each deidentified B-mode image to identify the cervical region for which Attenuation Coefficient, Backscatter Coefficient (dB), and Lizzi-Feleppa parameters were computed. To avoid selection bias, we mapped the entire cervix and used the mean values for each Quantitative Ultrasound biomarker. A biomedical engineer (AH, >10 years of experience) then analyzed the field of interest to compute the Attenuation Coefficient, Backscatter Coefficient (dB) and Lizzi-Feleppa parameters within the bandwidth 3.7–6.8 MHz by using a reference phantom approach.⁵⁵ Because Backscatter Coefficient (dB) and Lizzi-Feleppa Midband are directly related to one another, only Backscatter Coefficient (dB) is reported. The Shear Wave Speed values were provided on-screen and within a sonographer determined region of interest by the S2000 scanner without offline processing.

Statistical Analysis

Study data were collected and managed using Research Electronic Data Capture (REDCap)^56,57 hosted by UIC. Statistics were calculated using Excel (Microsoft Corporation, Excel for Mac 2022, Version 16.67, Redmond WA) and R (R Core Team, Version 4.3.2., 2023, R Foundation for Statistical Computing, Vienna, Austria) with the ‘lme4’ package.

A linear mixed-effects model (LMM) regression analysis was conducted to investigate the association between Quantitative Ultrasound features (Attenuation Coefficient, Backscatter Coefficient (dB), Shear Wave Speed, Lizzi-Feleppa Slope, and Lizzi-Feleppa Intercept) and the categorical independent variable, birth outcome (full-term or preterm) and the continuous variable, time after delivery (TAD). Linear mixed-effects models (LMM) were selected to address the repeated measures within participants at each visit, capture time trends, and to accommodate for missing data without excluding the participants with incomplete follow-up. We accounted for the repeated measures correlation by including random participant effects and sonographer effects in the LMMs. This approach mitigated the impact of missing data by leveraging the repeated measures from each participant over time. The standard errors were computed using 1000 bootstrapped samples providing a robust measure of variability that was less sensitive to the uneven sample sizes across time points. Two models were compared through likelihood ratio testing and Akaike Information Criteria (AIC).

The additive model with the lowest AIC value was selected and can be interpreted as:

$$y=({\beta }_0+\gamma +\delta )+{\beta }_{1}PTB+{\beta }_{2}TAD$$ (1)

where β₀ and β₁ and β₂ represent the model’s fixed-effect coefficients, and γ and δ are the uncorrelated random intercept for each participant and sonographer, respectively. FTB = 0 for full-term birth, PTB = 1 for preterm birth. Time after delivery (TAD) denotes a normalized variable calculated by dividing the days from delivery at the time of the scan visit by 365, expressed as a fraction of the year post-delivery. To review the model that was not selected refer to Tables A-1 and A-2 in the supplementary material.

RESULTS

The final analytical sample included 55 women; 41 women who delivered full-term and 14 women who delivered spontaneous preterm. Participant characteristics are summarized by group in Table 1. All preterm deliveries were spontaneous, none were medically induced. Most of the participants were multigravid; approximately half had a recurrent sPTB; and exactly half of the women who had a sPTB had a maternal history of sPTB. Most of the women who had a sPTB were Black or Hispanic (8 and 5 PTBs, respectively). There were no statistically significant (p>0.05) differences in mean maternal age, infant sex, or maternal family history of PTB between the two cohorts. The women who delivered preterm infants had lower birthweights than those who delivered full-term babies (p<0.01) and women with a history of prior PTB had a higher incidence of sPTB than women without a prior history (p<0.05).

Table 1.

Participant Characteristics, including demographic and gestational details are presented by birth outcome. Data are reported as counts and mean ± standard deviation, with range in parenthesis. The p-values are derived from two-tailed (0.05 significance level) t-test for quantitative values and two-sided Fisher’s Exact test (for category counts).

Characteristics	FTB cohort (n=41)	sPTB cohort (n=14)	p-value
Maternal Age (years)	29±6.6 (17–39)	29±4.1 (22–36)	0.77
Maternal Age count ≤ 24 yr	13	2	0.30
Maternal Age count 25–39 yr	28	12
GA at delivery (weeks)	(370/7-404/7)	(200/7-364/7)
Preterm type68
Extremely preterm: at or before 25 weeks	-	2
Very preterm: <32 weeks	-	2
Moderately preterm: between 32 – 34 weeks	-	2
Late preterm: between 34 – 36 completed weeks	-	8
Infant Sex
Male	19	8	0.55
Female	22	6
Infant Birthweight (g)	3279±403 (2280–4406)	1976±387 (562–2844)	0.0001***
Racial and Ethnic Categories
Asian	1	0	0.16
Black or African American	14	8
Hispanic or Latino	16	5
White	8	0
More than 1 race	0	1
Other	2	0
Maternal Family History of PTB
Yes	5	5	0.10
No	36	9
Previous PTB
Yes	9	8	0.021**
No	32	6
Prior sPTB counts
Nulliparous 0	16	6	0.049**
Multiparous 1	8	5
Multiparous 2	1	3
Prior PTB count + prior FTB count
Nulliparous 0	16	6	0.27
Multiparous 1	13	2
Multiparous 2	3	4
Multiparous 3	5	1
Multiparous 4	4	1
Last pregnancy outcome
Abortion	4	2	0.04**
Full-term birth	14	1
Induced preterm birth	0	0
Miscarriage	7	1
No prior pregnancy	12	4
Spontaneous preterm birth	4	6
Stillbirth	0	0
Progesterone
No	38	9	0.02**
Yes	3	5
17-Hydroxyprogesterone	2	3
Vaginal	1	2
Progesterone indication			1.0
Previous sPTB	2	3
Short cervix	1	2

Note: Asterisks (*) denote significance levels:

^*p<0.1

^**p<0.05

^***p<0.01

FTB: Full-term birth

sPTB: Spontaneous preterm birth

Table 2 summarizes the number of scans completed by 5 sonographers at each timepoint by cohort and Table 3 shows the distribution of participants based on visit frequency.

Table 2.

Counts of all participant scans completed at 6-weeks, 3-months, 6-moths, 9-months, and 12-months postpartum. The table displays the total number of scans completed, organized by visit in the year after delivery.

Group	Visit
	6-weeks n=54	3-months n=31	6-months n=26	9-months n=20	12-months n=24	Total
Full-term	40	24	20	18	19	121
Preterm	14	7	6	2	5	34

Note: One participant was unable to complete the 6-week scan and three 12-month full-term scans were omitted from the analysis due to a subsequent pregnancy.

Table 3.

Frequency distribution of the total number of visits completed by participants ts in the year after delivery organized by birth outcome.

Total Number of Visits Attended	Full-term	Preterm
1 Visit	15	6
2 Visits	4	1
3 Visits	3	4
4 Visits	6	1
5 Visits	13	2

Note: Table does not provide information about what specific visits were attended by participants.

Table 4 displays the linear mixed-effects model regression estimates to assess whether there were differences in postpartum cervical remodeling between women who delivered full-term and spontaneous preterm. Concerning the association between Quantitative Ultrasound biomarkers and birth outcome, the sPTB group was associated with significantly higher Backscatter Coefficient (dB) and Lizzi-Feleppa Intercept values compared to the full-term birth group (p<0.05). Additionally, the sPTB group exhibited higher coefficient values for Quantitative Ultrasound biomarkers Attenuation Coefficient, Shear Wave Speed, and Lizzi-Feleppa Slope, although these differences were marginal and not significantly different from the full-term birth group. Table 4 random effect variance estimates indicate that the inter-sonographer variance is virtually negligible compared to participant variance and noise (repeatability) variance.

Table 4.

Regression model estimates. The values in parentheses represent the p-values for the model coefficient tests. The outcome coefficient shown corresponds to the preterm group (preterm=1), indicating the average change in the outcome variable compared to the reference group (full-term birth group). Time after delivery was normalized as a continuous variable, expressed as a fraction of a year (calculated by dividing the number of days from delivery to the visit day by 365 days).

	AC	BSCdB	SWS	Slope	Intercept
Fixed Effect Coefficients
(Reg. Intercept)	0.89 *** (<0.001)	−24.00 *** (<0.001)	3.40 *** (<0.001)	2.43 *** (<0.001)	−36.80 *** (<0.001)
Outcome [Preterm]	0.10 (0.160)	2.08 ** (0.014)	0.25 (0.118)	0.08 (0.654)	1.72 ** (0.010)
Days from delivery	0.16 *** (<0.001)	2.16 *** (<0.001)	−0.19 ** (0.014)	0.17 ** (0.020)	1.39 *** (<0.001)
Random Effect Components of Variance
σe2 (Noise component)	0.07	10.36	0.42	0.46	5.87
σID2 (Participant component)	0.04	6.88	0.26	0.30	4.57
σ2 Sonographer (Sonographer component)	0.00	0.13	0.01	0.01	0.68
ICC	0.40	0.39	0.40	0.47	0.40
Nid	55	55	55	55	55
Model Fit Statistics
Deviance	490	8021	3145	3296	7174
log-Likelihood	−252	−4010	−1577	−1652	−3615

Note: Asterisks (*) denote significance levels:

^*p<0.1

^**p<0.05

^***p<0.01

ICC: Intraclass correlation

AC: Attenuation Coefficient in units of dB/cm-MHz

BSCdB: Backscatter Coefficient in units of 1/sr-cm that were log-transformed [i.e., 10 log₁₀(BSC)]

SWS: Shear Wave Speed in units of m/s

Slope: Lizzi-Feleppa Slope in units of dB/MHz

Intercept: Lizzi-Feleppa Intercept in units of dB

In addition, Table 4 shows the linear mixed-effects model estimates used to evaluate cervical remodeling over 12-months post-delivery. Concerning the association between Quantitative Ultrasound biomarkers and the time after delivery, as the time from delivery increased, there was a significant positive relationship among Quantitative Ultrasound biomarkers Attenuation Coefficient, Backscatter Coefficient, Lizzi-Feleppa Slope, and Lizzi-Feleppa Intercept (p<0.05). Therefore, as more time passed after delivery, these biomarkers had higher values. In contrast, there was a significant negative association between Shear Wave Speed and the time after delivery (p<0.05), indicating that as the time after delivery increased, Shear Wave Speed values decreased. Figure 2 shows the main effect plot of cervical remodeling over time in relation to each Quantitative Ultrasound biomarker across the two groups.

Figure 2.

Main effect plots show the change in each Quantitative Ultrasound biomarker over 12-months post-delivery across the full-term and sPTB groups. The shaded lines represent the 95% confidence interval.

Note: Black tick counts indicate the number of participants completing the postpartum scan visit. Only Backscatter Coefficient (BSC) and Lizzi-Feleppa Intercept are significantly different between the full-term and preterm birth groups.

DISCUSSION

In the first 12 months post-delivery, significant differences were observed in all Quantitative Ultrasound biomarkers and trajectories for the full-term and spontaneous preterm birth groups. It is not clear why these differences exist and their clinical significance. Our findings align with study³⁵ results that suggest postpartum cervical remodeling is an ongoing process in the year post delivery.

This is the first study to noninvasively identify differences in postpartum cervical remodeling between women who delivered full-term and spontaneous preterm with Quantitative Ultrasound. In this study, Quantitative Ultrasound serves as a proxy for tissue remodeling, specifically collagen remodeling in the cervix. Delivering preterm was associated with markedly higher Backscatter Coefficient (dB) and Lizzi-Feleppa Intercept values in comparison to the full-term birth group. The higher Backscatter Coefficient (dB) and Lizzi-Feleppa Intercept values in the cervical tissue of the sPTB group suggest a potential increase in the presence of scattering elements (i.e., number or density of structures within the cervical tissue that scatter ultrasound waves) and/or alterations in how the tissue interacts with the sound waves. Based on prior research, there is evidence to suggest that ultrasound scattering is related to cervical collagen content and organization.^19,58 In one of our previous studies evaluating cervical remodeling during pregnancy in the rat cervix,¹⁹ we reported that as the pregnancy progressed the concentration of scatterers decreased, reflective of increased collagen content and decreased collagen concentration. In cervical tissue, this change was observed as disorganization of the collagen fibers and an increase in the amount of space between the fibers, consistent with cervical tissue morphology in the rat cervix during remodeling.^59,60 Consistent with these studies, we hypothesize that women who had a sPTB may have different collagen composition that interacts with the ultrasound waves, producing a different pattern of scattering. Our findings suggest that there may be different mechanisms of postpartum cervical recovery between spontaneous preterm and full-term birth remodeling, possibly due to differences in collagen morphology. However, these hypotheses are yet to be tested. While our findings do not pinpoint the specific components in collagen that act as effective scattering elements, we hypothesize that alterations in the organization of the collagen fibers, the composition of the extracellular matrix (including proteoglycans), and the inflammatory response may alter cellular and extracellular properties that may contribute to higher values in the sPTB groups. Evidence from animal and limited human studies support our hypotheses.^{19,25–27,41,61,62} Further research is needed to determine the implications of higher Backscatter Coefficient and Intercept values in postpartum cervical tissue and to fully understand the relationship between collagen organization and the ultrasound biomarkers. Nevertheless, these results are encouraging because they offer direction for future studies. Future studies could focus on determining the reliability of Backscatter Coefficient and Lizzi-Feleppa Intercept in a larger sample with more defined subgroups (full-term, spontaneous preterm, and recurrent spontaneous preterm) to provide a more comprehensive characterization of postpartum cervical remodeling and its role in pretem and recurrent preterm birth.

In this study, we determined that Quantitative Ultrasound biomarkers were able to detect differences in cervical remodeling between women who delivered full-term and spontaneous preterm. Our findings were also supported by studies^36,37 that reported there were differences in collagen concentration post-delivery between women with cervical insufficiency that experienced a late spontaneous abortion or PTB and women without cervical insufficiency.

Clinical and Research Implications

The results of this study suggest that there may be cervical remodeling differences between full-term and preterm deliveries and that using Quantitative Ultrasound biomarkers may enhance our understanding of cervical remodeling in the postpartum period. Being able to identify women at higher risk of experiencing another preterm birth would enable clinicians to offer personalized support, including preconception counseling and family planning services so that women can make informed decisions about their reproductive options before they are pregnant again. Additionally, monitoring cervical remodeling with Quantitative Ultrasound biomarkers has the potential to provide insights into the changes occurring in cervical tissue in response to the interventions that are prescribed, thus potentially influencing the choice of treatments offered or modifying the treatments based on tissue response. Therefore, continued research into postpartum cervical microstructure is critical for the development of interventions that lead to precision health approaches that predict and prevent recurrent sPTB.

Strengths and Limitations

A strength of this study was that the ultrasonic radiofrequency data can be analyzed and reanalyzed without additional patient data collection. This study was also strengthened by the preliminary pilot animal and human Quantitative Ultrasound findings that determined that Quantitative Ultrasound biomarkers are capable of detecting collagen microstructure changes in the cervix in both animals and humans.^19,30,41

Limitations of the study include the small sample size and limited number of sPTB participants. Another limitation was that although the American College of Obstetricians and Gynecologists (ACOG) has released guidelines supporting ongoing, individualized postpartum care; routine care usually only includes one visit at approximately 6 weeks after delivery.⁶³ This standard of care does not include transvaginal scans to assess cervical healing and women are usually not routinely followed past 12-weeks postpartum. Therefore, the transvaginal scans and timepoints selected for this study represent a novel approach to postpartum care which would require further research and if successful, policy efforts to implement into routine clinical practice. Additionally, only a few commercial diagnostic ultrasound scanners (e.g., Siemens Acuson Sequoia and GE Logiq E10) currently provide access to the raw radiofrequency data required to implement the Quantitative Ultrasound technology reported in this study.^64–67 This greatly limits the opportunity to use Quantitative ultrasound technology for research and ultimately impedes the development of targeted interventions for recurrent sPTB. This limitation could be alleviated if/when more ultrasound manufacturers start to provide radiofrequency data acquisition capabilities on their clinical machines.

Conclusions

Quantitative Ultrasound is a noninvasive technology that can improve our understanding of postpartum cervical remodeling at the tissue level. Our results suggest that postpartum cervical remodeling differs between women who deliver full-term and spontaneous preterm. These findings lay the groundwork for future research that may provide a scientific basis supporting the use of Quantitative Ultrasound biomarkers to investigate cervical tissue microstructure in women at high risk for recurrent spontaneous preterm birth.

Data Availability Statement

Data are available on reasonable request from the corresponding author.