The impact of delayed evacuation on the quality of human fetal tissue

TL;DR

This study examines how delaying fetal tissue removal affects tissue quality, finding that warm ischemia after circulation stoppage causes significant degradation.

Contribution

The study is the first to systematically assess the impact of delayed evacuation and digoxin on human fetal tissue quality.

Findings

Delayed evacuation leads to structural disruption, reduced cell proliferation, and increased apoptosis in fetal tissues.

RNA integrity is significantly reduced when tissues are exposed to warm ischemia and digoxin.

DNA remains stable even after delayed evacuation, making it suitable for PCR analysis.

Abstract

Agents such as digoxin and potassium chloride are frequently administered to arrest fetal circulation prior to termination procedures. Although this practice is common, the impact of delayed evacuation on human fetal tissue quality has not been systematically assessed. High-quality tissue is essential for accurate diagnosis of fetal abnormalities and for biomedical research.

To evaluate the effects of induced fetal demise and delayed evacuation on human fetal tissue quality across multiple organ systems, and to determine whether digoxin independently contributes to tissue degradation.

Tissue was collected from second-trimester dilation and evacuation procedures performed either after induced demise with digoxin or potassium chloride followed by evacuation 19–25 hours later, or without prior induction (immediate evacuation). Brain, lung, kidney, liver, and muscle were evaluated for morphology, proliferation, apoptosis, cell viability, culture success, and nucleic acid integrity. To model in utero conditions, tissues were incubated at 37 °C or 4 °C for up to 18 hours, with and without digoxin.

Delayed evacuation was associated with structural disruption, reduced proliferation, increased apoptosis, diminished fibroblast viability, and reduced RNA integrity compared to immediate evacuation. DNA integrity remained robust, with all samples suitable for PCR. Digoxin alone did not significantly alter RNA quality; however, when combined with prolonged incubation at 37 °C, RIN and DV200 values were significantly reduced across multiple tissues.

Warm ischemia following circulatory arrest is the principal driver of tissue degradation, with RNA and cell viability most affected and DNA relatively preserved. Digoxin alone has minimal effect but accelerates degradation under warm conditions. Minimizing the interval between circulatory arrest and evacuation, and avoiding prolonged exposure to body temperature, may preserve tissue integrity and improve the diagnostic and research utility of HFT.

Figures

Histological Analyses of Brain, Kidney and Lung from Specimens after Delayed (DE) and Immediate (IE) Evacuation.A. H&E staining shows an increased presence of small/condensed nuclei with cytoplasmic swelling and vacuolization, in the delayed evacuation panel. B. More pronounced TUNEL staining (brown), indicating greater apoptosis in the DE panel. C. A few or absence of Ki-67 staining (brown), indicating decreased cell division in the delayed evacuation panel. Scale bar = 20μm.

Brain Cell Viability.Mean viability (black bar) for brain tissue collected from DE specimens (black circle, n = 3) and IE specimens (white circle), incubated at 37°C for 0, 6 (n = 7), 12, 18, and 24 (n = 6) hours prior to tissue dissociation for cell culture. Percent viability was calculated as the number of trypan blue negative cells divided by the total number of cells. Student’s t-test, *P < 0.05, P < 0.001.

RNA Quality with Delayed Evacuation Compared to ex vivo Incubation at 37°C.A. RIN scores for brain, kidney, liver, lung, and muscle from DE (n = 7) and IE (n = 5) specimens incubated at 37°C for 0, 6, or 18 hours. Boxplots display the median (horizontal line), the interquartile range (box), and whiskers (extending to ±1.5 × the IQR) with individual replicate data points overlaid. Across all tissues, IE samples exhibited significantly higher RIN scores than DE at baseline, with progressive degradation over time. B. DV200 scores for the same tissues. Boxplots illustrate the distribution of DV200 values, with IE samples retaining higher DV200 values than DE; An apparent decline is observed after 18 hours. Statistical significance was determined by one-way ANOVA (P < 0.05 for all tissues) and Student’s t-test; *P < 0.05, P < 0.005.

RNA Quality with ex vivo Digoxin Treatment.A. RIN scores for brain, kidney, liver, lung, and muscle from IE specimens (n = 4): untreated, and digoxin-treated samples incubated for 18 hours at 4°C or 37°C. Boxplots indicate the median (horizontal line), interquartile range (box), and whiskers (±1.5 × IQR), with individual points representing replicate data. RIN values remained high in untreated and 4°C samples but declined significantly at 37°C across tissues. B. DV200 scores for the same samples. DV200 values were largely preserved in untreated and 4°C groups but dropped notably after 37°C incubation. One-way ANOVA (P values shown above groups) followed by Student’s t-test; not significant (P > 0.05).

Full text

Introduction

Evaluation of fetal tissue can be critical for determining a specific biomedical diagnosis in pregnancies affected by fetal abnormalities. A specific diagnosis is essential for accurate genetic counseling of recurrence risk and outcomes for future pregnancies. The historical standard for fetal diagnosis is an autopsy by a fetal pathologist, which can include radiological examinations, gross and microscopic tissue evaluation (including immunohistochemical staining), cellular assays, and select biochemical testing [1,2]. Increasingly, fetal DNA is used for microarray testing to detect pathogenic chromosome copy number variants, as well as single gene, gene panel, and exome or whole genome sequencing (WGS) approaches to identify causative DNA variants [2–5].

In parallel with clinical applications, HFT obtained after pregnancy termination has played a vital role in research advances that have greatly improved health care over decades. Human fetal-derived cells used for viral culture were essential for generating vaccines to poliovirus, varicella, rubella, measles and other pathogens [6–10]. Vaccines created using the WI-38 fetal lung cell line alone are estimated to have averted or ameliorated around 4.5 billion cases of poliomyelitis, measles, mumps, rubella, varicella, herpes zoster, adenovirus, rabies and Hepatitis A infection worldwide from 1960 to 2015 [11]. The human embryonic kidney293 (HEK293) fetal cell line has been widely used in a number of studies, including recombinant protein production, gene therapy, and drug development, as well as more recently in the development of RNA-based vaccines for COVID-19 [12–14]. HFT provision for the Human Lung Cell Atlas has also facilitated insights into the precise pathogenesis of COVID-19 infection [15]. Furthermore, innumerable benefits of HFT provision to many other biomedical research efforts exist including developing immunotherapies for cancer, stem cell biology, tissue and organ cell regeneration, human disease modeling in animal models and characterizing birth defects, among others [16].

In the United States, most second trimester pregnancy terminations are completed by dilation and evacuation (D&E) [17,18]. D&E procedures are often preceded by the use of digoxin, potassium chloride (KCl), and other agents to arrest fetal circulation [19]. This approach is used for both termination of pregnancies (TOP) with structurally normal fetuses and those with fetal anatomic anomalies, genetic disorders, infections, or teratogenic exposures [19]. While no evidence-based medical benefits of this practice have been demonstrated, [20] some clinicians perceive increased procedural ease and safety [21]. In addition, institution, provider, and/or patient preference may play a role. In the U.S., clinician concern for legal ramifications of the Partial Birth Abortion Ban Act of 2003 is also influential [19,22].

Methods of inducing fetal demise include injecting the agent into the fetus, umbilical cord, or amniotic fluid, typically followed by D&E procedure or labor induction [19]. While concerns have been raised that the method of termination may adversely impact the utility of fetal tissues, [23] no studies have systematically evaluated the effects of agents that arrest fetal circulation or the timing of D&E after injection on HFT quality. To address this critical knowledge gap, the goal of this study was to evaluate HFT quality using a variety of parameters that evaluate utility of these tissues for diagnostic and research purposes.

Materials and methods

Collection of human fetal tissue

Specimens were obtained from several outpatient reproductive health clinics in the Seattle area between March 18^th^, 2014 and December 7^th^, 2018. All specimens were collected from pregnant participants who chose to donate HFT during a D&E procedure. Second-trimester pregnancy terminations comprised most donations, with four out of 44 specimens being early third-trimester. Written informed consent was obtained from all participants under a protocol approved by the University of Washington Institutional Review Board (STUDY00000380). Participants were consented broadly through the Birth Defect Research Laboratory (BDRL) for potential use of donated tissue in biomedical research only after consent for the procedure was obtained. Clinics followed varying protocols concerning the gestational duration at which digoxin or KCl was administered, with most using these agents after 18–20 weeks post-fertilization.

HFT from 44 specimens were included in this study (S1 Table). HFT was collected from D&E procedures performed approximately 24 hours (19–25 hours) after the induction of fetal demise using digoxin (intra-fetal or intra-amniotic) or KCl (intra-cardiac or intra-funic), consistent with standard care practices in Washington State. Injections coincided with the placement of cervical dilators per clinic protocols. Providers confirmed fetal demise via ultrasound prior to the D&E. The research team did not receive full ultrasound reports; limited ultrasound measurements (e.g., crown–rump length, biparietal diameter, or femur length) were provided separately and used only as secondary confirmation of gestational age if post-mortem foot length was unavailable. Specimens were also collected from D&E procedures completed without prior induction of fetal demise. Tissues that were not processed immediately were stored at 4°C during transport back to the university. Upon arrival to the laboratory, all tissues were briefly washed in Hanks’ balanced salt solution (HBSS) prior to processing for downstream applications. Participant data was de-identified and assigned a study ID number to be analyzed anonymously. The gestational duration of the fetus was measured from fetal foot length and reported as the number of days post-fertilization [24]. We compared tissues from specimens evacuated 24 hours after digoxin or KCl injection (delayed evacuation) to tissues from specimens evacuated without prior injection (immediate evacuation). This was intended to assess whether the timing of evacuation after induced demise alters tissue integrity across key clinical and research metrics.

Target organs (brain, kidney, liver, lung, and muscle) were selected both to reflect the most frequently requested tissues from collaborating researchers and to represent a range of functionally and developmentally distinct organ systems. They were evaluated in parallel to determine whether ischemic sensitivity varies across tissue types, with implications for both diagnostic pathology and downstream research use. Not all specimens were sufficiently intact to provide all target organs, so sample sizes varied by tissue and assay as described in the Results and figure legends.

To assess the feasibility of maintaining viable cell populations, fibroblast cultures from skin and astrocyte cultures from brain were established, as these cell types are commonly used in diagnostic assays and functional research applications that depend on long-term culture [25,26]. The histochemistry (TUNEL) and immunohistochemistry (Ki67) assessments summarized in S2 Table were the only analyses performed under blinded conditions, with investigators unaware of specimen group assignments. Due to limited sample availability, the number of specimens for both immediate and delayed evacuations was insufficient for consistent inclusion across all analyses: histological evaluations (3 delayed and 3 immediate evacuations), cell viability tests (3 delayed and 7 immediate evacuations), cell culture tests (11 delayed and 9 immediate evacuations), and nucleic acid tests (8 delayed and 2 immediate evacuation specimens).

Given these constraints, controlled incubation experiments were designed to further define the impact of ischemia and digoxin exposure on tissue quality. Warm ischemia was modeled to evaluate its impact on tissue quality in the absence of induced demise. Tissues from two immediate evacuation specimens were incubated at 37°C in Gibco Hibernate-E for 6 and 18 hours. These incubated controls were analyzed as described below and compared with delayed evacuation tissues and non-incubated immediate evacuation tissues. To investigate the potential effects of digoxin on tissue quality, we incubated tissues from four immediate evacuation specimens in Gibco Hibernate-E with and without digoxin (1 mg/500 mL media) at both 37°C and 4°C for 18 hours. This approach allowed us to distinguish the effects of digoxin from those of ischemia and delayed evacuation.

Histological evaluation & cell culture

Brain, kidney, liver, lung, and muscle tissues were fixed in 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS) before processing and embedding in paraffin. Serial 5 μm sections were mounted on glass slides. Cellular morphology was assessed using hematoxylin and eosin (H&E) staining. Cell shape, cytoplasmic integrity, and cytoskeletal organization were evaluated by immunohistochemistry using a Vimentin antibody (DAKO, Cat #M0725, 1:2,000) [27]. Apoptosis was assessed using the TUNEL assay (Millipore, Cat #S7101) with methyl green as a counterstain. Cell proliferation was evaluated using a Ki67 antibody (DAKO, Cat #M7240, 1:200).

To establish fibroblast cultures, fetal skin was digested with Collagenase Type 2 (Worthington, Cat #LS004174) in Minimum Essential Medium α (Gibco, Cat #12571–063) for 1–2 hours, according to departmental protocols, at 37°C before plating in Solid Tissue Media (Minimum Essential Medium α, 10% fetal bovine serum, 1% penicillin/streptomycin, 0.2% Normocin). Astrocytes were cultured according to previously established protocols [25]. Cell viability was estimated based on the exclusion of the vital stain, 0.4% trypan blue, prior to the initial plating of the primary cell culture.

Nucleic acid extraction and quality analysis

Tissue samples (brain, kidney, liver, lung, muscle) designated for DNA analysis were stored at −80°C before genomic DNA extraction using the Maxwell 16 system (Promega Corporation). Agarose gel electrophoresis confirmed the presence of high molecular weight DNA. DNA integrity was assessed using an Agilent 2200 Tape Station and expressed as the DNA Integrity Number (DIN). To evaluate the suitability of genomic DNA for standard molecular diagnostic applications, polymerase chain reaction (PCR) was performed using primers for the SRY and AMELX genes: AMXY-F: CTGATGGTTGGCCTCAAGCCTGTG, AMXY-R: TAAAGAGATTCATTAACTTGACTG, SRY-F: AGTGTGAAACGGGAGAAAAC, SRY-R: TACAACCTGTTGTCCAGTTG [28].

For RNA isolation, tissue samples (brain, kidney, liver, lung, muscle) were stored in RNAlater (Qiagen, Cat #76104) according to the manufacturer’s instructions before total RNA extraction using the Maxwell 16 system (Promega Corporation). RNA quality was assessed using an Agilent 2100 Bioanalyzer to determine the RNA Integrity Number (RIN) [29] and the DV200 (percentage of RNA fragments >200 nucleotides) [30].

Statistical analysis

Statistical analyses were conducted in RStudio (version 2023.09.0 + 463, Posit Software, PBC). Comparisons of cell viability between DE and IE groups were performed using two-sample t-tests. Linear regression was used to model the relationship between incubation duration and cell viability. RNA integrity (RIN) and DV200 values were compared between DE and IE groups and across incubation conditions using two-sample t-tests and one-way ANOVA with Tukey’s HSD post hoc tests. All p-values were two-sided, and values < 0.05 were considered statistically significant. Graphs and boxplots were generated using the ggplot2 package in R.

Because RNA is highly susceptible to degradation and directly impacts downstream applications like transcriptomics, RNA integrity is more frequently quantified in tissue quality studies [29,30]. DNA is typically more stable, and limited sample sizes often precludes formal statistical analyses of DIN [31]. For this reason, DIN values were determined only for a subset of specimens undergoing PCR and reported descriptively (S1 Fig and S1 Table) without formal statistical comparison.

Results

We conducted a comparative analysis of tissue samples from 18 specimens subjected to delayed evacuation (DE) after either digoxin or potassium chloride (KCl) injections, 19–25 hours prior, against samples from 26 specimens undergoing immediate evacuation (IE) without prior injections (Tables 1 and S1). Among these, four specimens exhibited congenital anomalies. The average post-conception age, as indicated by fetal foot length, was significantly higher in the DE group (145 days) compared to the IE group (121 days) (S1 Table). Digoxin and KCl doses ranged from 0.5 mg to 1.5 mg, with two cases receiving both agents. Most DE specimens displayed noticeable degradation (softening and discoloration), complicating the blinding process for analyses.

Table 1: Specimen characteristics.

Histological evaluation

We performed histological evaluations of brain, kidney, and lung from three DE specimens compared to three IE specimens (S1 Table). By H&E staining, key indicators of disrupted cellular morphology including cytoplasmic swelling, vacuolization, and small, condensed nuclei that were markedly more prevalent in DE tissues (Fig 1A). Similarly, TUNEL staining was more pronounced in DE tissues, indicating severe apoptosis (Fig 1B), and Ki67 staining was nearly absent in DE tissues (Fig 1C), indicating decreased cell division and/or loss of Ki67 antigen due to degradation (Fig 1C and S2 Table). Comparisons of staining estimates for both markers at 0-hour time-point can be compared by viewing DE to IE (S2 Table).

Histological Analyses of Brain, Kidney and Lung from Specimens after Delayed (DE) and Immediate (IE) Evacuation.A. H&E staining shows an increased presence of small/condensed nuclei with cytoplasmic swelling and vacuolization, in the delayed evacuation panel. B. More pronounced TUNEL staining (brown), indicating greater apoptosis in the DE panel. C. A few or absence of Ki-67 staining (brown), indicating decreased cell division in the delayed evacuation panel. Scale bar = 20μm.

To simulate the effect of intrauterine degradation by exposure to prolonged body temperature incubation on tissue quality, we incubated IE tissues at 37°C for 6 and 18 hours (S2 Table). At 6 hours, TUNEL and Ki67 staining were mildly altered compared to the 0-hour time point. At 18 hours, the incubated tissues appeared markedly degraded compared to the 0 and 6-hour time points, showed substantial TUNEL staining and lacked Ki67 staining indicating high levels of apoptosis and a complete absence of cell proliferation. The 18-hour tissues also appeared visibly more degraded than typical delayed evacuation tissues.

Cell culture

To evaluate the ability of tissues to continue growing in culture, we sampled skin from available sites on the integument of specimens and attempted to establish fibroblast cultures. Fibroblast cultures were successfully established in 3/11 DE specimens compared to 9/9 from IE specimens. The time interval between injection of the agent and the D&E did not clearly differ between the specimens with successful and unsuccessful fibroblast cultures. In addition, the successful cultures came from specimens with and without anomalies, injected with both digoxin and KCl, and injected at different sites (intra-amniotic and intra-fetal).

We compared cell viability in three DE specimens to seven IE specimens and assayed cell viability using trypan blue staining (Fig 2). The mean percent cell viability by trypan blue staining for the DE specimens was 22% (n = 3, SD = 13%) compared to 38% (n = 7, SD = 6.7%) for IE specimens. Incubating IE brain tissue at body temperature for 6 hours lowered the viability to 16% (n = 7, SD = 5.5). Incubating for 12, 18, and 24 hours further reduced viability to 8.8% (n = 6, SD = 5.7), 5.5% (n = 6, SD = 4.5), and 1.8% (n = 6, SD = 1.9) respectively. A two-sample t-test was performed to compare percent cell viability in DE tissue and IE tissue with no incubation. There was a significant difference in cell viability between DE tissue and IE tissue (p = 0.033). Simple linear regression was used to test if hours incubated significantly predicted cell viability. The fitted regression model was Percent cell viability = −1.39(hours incubated) + 30.6. The overall regression was statistically significant (R^2^ = 0.7261, F (1,30) =79.5, p < 0.001). It was found that hours incubated significantly predicted cell viability (β = −1.39, p < 0.001).

*Brain Cell Viability.Mean viability (black bar) for brain tissue collected from DE specimens (black circle, n = 3) and IE specimens (white circle), incubated at 37°C for 0, 6 (n = 7), 12, 18, and 24 (n = 6) hours prior to tissue dissociation for cell culture. Percent viability was calculated as the number of trypan blue negative cells divided by the total number of cells. Student’s t-test, *P < 0.05, *P < 0.001.

Despite the presence of viable cells based on trypan blue staining, no cells adhered to plates or grew from DE specimens using our standard astrocyte cell culture protocol [25]. In contrast, astrocyte cells adhered and grew from all 11 IE brain samples. In 6/7 IE brain samples incubated for 6 hours adhered and generated cell lines. By 12 hours incubation, 2 out of the remaining 6 grew, by 18 hours 1/6 grew, and by 24 hours, no cells adhered or grew.

Nucleic acid quality: DNA & RNA

To determine DNA quality from specimens, we isolated high molecular weight DNA from all tissues, calculated the DNA Integrity Number (DIN), and performed PCR. Based on agarose gel electrophoresis, we were able to isolate high molecular weight DNA from all tissues. However, the quality differed by organ and between DE and IE tissues. DNA extracted from brain and muscle demonstrated high molecular weight DNA with minimal smearing, whereas DNA extracted from lung, kidney, and liver showed increased smearing. While smearing may suggest DNA degradation, it can also result from factors such as RNA, protein, or salt contamination, gel preparation, well loading, or sample handling. We cannot rule out these factors entirely, though steps were taken to minimize contamination during processing.

Similarly, while DNA Integrity Numbers (DIN) were acceptable across all tissues, DE tissues generally had lower DIN scores than IE tissues (S1 Fig and S1 Table). These findings indicate that while overall DNA quality remained sufficient for downstream applications such as PCR, delayed evacuation tissues may exhibit subtle compromises in DNA integrity compared to immediate evacuation tissues.

We performed PCR for sex determination using genomic DNA extracted from kidney in 8 delayed evacuation specimens and 2 immediate evacuation specimens. All 10 samples amplified appropriately, and the product sizes (977 base pairs for the X chromosome; 790 and 358 base pairs for the Y chromosome) were consistent with the external morphological sex assignment (S1 Fig).

To determine the impact of DE on RNA quality, we isolated RNA from each specimen and calculated RIN scores. RIN scores for delayed evacuation tissues were all less than 6 (range 3.5–5.6) whereas Immediate evacuation tissues consistently yielded high quality RNA, with all RIN scores >8 (Fig 3A and S1 Table). Using a two-sample t-test, we compared RIN scores in DE and IE tissues that were not incubated. We found significantly higher RIN scores in DE tissues for brain (p = 0.0016), kidney (p < 0.001), liver (p < 0.001), lung (p < 0.001), and muscle (p = 0.0020). Next, a one-way ANOVA evaluated the effect of different incubation times (0h, 6h, 18h) on RIN scores for each organ. We observed significant changes in brain (p = 0.012), kidney (p = 0.0027), liver (p < 0.001), lung (p = 0.0011), and muscle (p < 0.001). Tukey’s HSD Test showed that RIN scores generally decreased between 0h and 18h, and in many tissues between 6h and 18h as well. Brain tissue began to degrade by 6 hours at 37°C, whereas muscle was the most resistant.

*RNA Quality with Delayed Evacuation Compared to ex vivo Incubation at 37°C.A. RIN scores for brain, kidney, liver, lung, and muscle from DE (n = 7) and IE (n = 5) specimens incubated at 37°C for 0, 6, or 18 hours. Boxplots display the median (horizontal line), the interquartile range (box), and whiskers (extending to ±1.5 × the IQR) with individual replicate data points overlaid. Across all tissues, IE samples exhibited significantly higher RIN scores than DE at baseline, with progressive degradation over time. B. DV200 scores for the same tissues. Boxplots illustrate the distribution of DV200 values, with IE samples retaining higher DV200 values than DE; An apparent decline is observed after 18 hours. Statistical significance was determined by one-way ANOVA (P < 0.05 for all tissues) and Student’s t-test; *P < 0.05, *P < 0.005.

DV200 measures the percentage of RNA fragments longer than 200 nucleotides. From the recorded specimens, DV200 ranged from 73–91%, indicating moderate or better RNA quality (Fig 3B). A two-sample t-test comparing DE and IE tissues without incubation revealed significantly higher DV200 values in DE for kidney (p < 0.001), liver (p = 0.0010), lung (p < 0.001), and muscle (p = 0.019), but not for brain (p = 0.073). A one-way ANOVA was then used to determine how incubation time (0h, 6h, 18h) affected DV200. Significant differences were found for brain (p = 0.021), kidney (p < 0.001), lung (p = 0.011), and muscle (p = 0.0014), but not for liver (p = 0.16). Tukey’s HSD showed that DV200 values dropped between 0h and 18h in brain, kidney, lung, and muscle, and between 6h and 18h in kidney, lung, and muscle. For tissues that were not incubated, a two-sample t-test found no significant difference in RIN scores between samples with and without digoxin (brain: p = 0.96, kidney: p = 0.50, liver: p = 0.16, lung: p = 0.78, muscle: p = 0.50). However, when digoxin-treated tissues were incubated for 18 hours at either 4°C or 37°C, RIN scores were significantly lower at 37°C for brain (p = 0.015), kidney (p = 0.0012), liver (p = 0.0092), lung (p = 0.025), and muscle (p = 0.018).

Similarly, in non-incubated tissues, there was no significant difference in DV200 values between digoxin-treated and untreated samples (brain: p = 0.76, kidney: p = 0.53, liver: p = 0.22, lung: p = 0.76, muscle: p = 0.94). But after 18 hours of incubation at 4°C or 37°C, DV200 was significantly lower at 37°C for brain (p = 0.021), kidney (p = 0.037), and liver (p = 0.044). No significant difference was found for lung (p = 0.23) or muscle (p = 0.077) between these temperatures.

Discussion

Principal findings

While fetal pathologists commonly observe that tissues from fetuses treated with digoxin, KCl, or similar agents are often markedly autolyzed, this phenomenon has not been systematically studied. This study does not assess clinical diagnostic outcomes but instead focuses on evaluating tissue integrity under different conditions. Here, we demonstrate that induced fetal demise followed by delayed evacuation is associated with greater cellular damage and apoptosis, reduced cell proliferation, and poorer nucleic acid quality. Incubating fetal tissues at 37°C (a proxy for delayed evacuation) causes rapid degradation, while digoxin treatment alone has minimal effects regarding RNA integrity. Comparison of tissues with and without ex vivo digoxin exposure for 18 hours at 4°C and 37°C showed no significant effect of digoxin on RNA quality, as measured by RIN and DV200 scores (Figs 4A-4B), which are two widely used metrics for RNA integrity and fragment distribution [29,30]. These findings suggest that prolonged incubation at body temperature without fetal circulation (“warm ischemia”) is the primary driver of the poorer tissue quality observed in terminated fetuses. Among the metrics assessed, cell viability and RNA integrity appear to be the most sensitive to ischemic conditions, whereas DNA integrity is relatively robust, with tissue histology and immunohistochemistry/staining falling between these extremes. This observation is consistent with prior studies showing that DNA is less vulnerable than RNA to degradation during tissue processing [31].

RNA Quality with ex vivo Digoxin Treatment.A. RIN scores for brain, kidney, liver, lung, and muscle from IE specimens (n = 4): untreated, and digoxin-treated samples incubated for 18 hours at 4°C or 37°C. Boxplots indicate the median (horizontal line), interquartile range (box), and whiskers (±1.5 × IQR), with individual points representing replicate data. RIN values remained high in untreated and 4°C samples but declined significantly at 37°C across tissues. B. DV200 scores for the same samples. DV200 values were largely preserved in untreated and 4°C groups but dropped notably after 37°C incubation. One-way ANOVA (P values shown above groups) followed by Student’s t-test; not significant (P > 0.05).

As with any medical intervention, healthcare providers and patients must weigh the risks and benefits of inducing fetal demise prior to D&E. Studies addressing the safety of digoxin for induction of fetal demise prior to termination are mixed. One study of 52 individuals randomized to two different doses of digoxin did not report adverse effects or an increase in nausea after digoxin administration [32]. However, a large retrospective cohort study comparing patients with and without induced demise demonstrated a three-fold increased risk of complications (vomiting, intrauterine infection, and unplanned delivery) [33]. Rare but serious risks to maternal health cannot be discounted and include hyperkalemic paralysis from digoxin [34] and cardiac arrest from KCl [35]. No clear evidence-based medical benefits of induced fetal demise prior to D&E or induction have been reported [21]. Patient experiences with induced fetal demise prior to D&E are complex; in one study, women reported both emotional discomfort and reassurance related to digoxin use [36].

Clinical and research implications

Different termination methods affect tissue quality in different ways. For instance, D&E causes physical disruption of the fetus, obscuring many gross anatomical features, but leaving histological features, cell viability, and nucleic acids largely intact. In contrast, induction termination results in largely intact gross anatomy. Induced fetal demise, which can occur prior to D&E or induction termination, variably impacts histology but severely impacts cell viability and RNA quality [29,37]. This is because RNA is inherently less stable than DNA in postmortem and ischemic contexts, with enzymatic activity and elevated temperatures accelerating degradation [38]. Such degradation has direct implications for traditional pathological evaluations, notably autopsy and its associated procedures for accurately diagnosing fetuses with abnormalities and impacting the viability of skin fibroblasts for any ancillary testing [2]. As a result, delayed evacuation after induced demise has the potential to impair the ability of clinicians to accurately diagnose fetuses with anomalies, thereby limiting prognostic and recurrence risk counseling, and determining future reproductive options for the patient. S1 Table summarizes the outcome of the choice of termination procedure and prior induction of demise on tissue utility for both clinical and research purposes.

Although fetal DNA is relatively resistant to degradation [31], the increasing identification of DNA variants of uncertain significance (VUS) through clinical genetic testing often requires additional laboratory evaluations to resolve whether the variants are causal or not [38]. For instance, DNA variants may affect RNA expression or splicing, which can be directly tested in RNA extracted from living cells, such as fibroblasts. RNA sequencing is emerging as a primary diagnostic technique, making the vulnerability of RNA especially concerning [39,40]. Even moderate RNA degradation can distort RNA-seq results, leading to inaccurate estimates of transcript abundance [37]. Further, direct DNA sequencing of multiple affected tissues is emerging as an effective diagnostic strategy for identifying mosaic disorders, where detection of pathogenic DNA variants is restricted to a subset of tissues [41].

Similar concerns arise for the utility of HFT in research. For many current cellular and developmental biology applications, viable cells are an essential starting material [42]. Maintaining a mitotically active cell line, such as fibroblasts, is essential for functional work to determine how DNA variants impact RNA expression or protein translation [43,44]. Use of HFT for research has been the focus of intense scrutiny by the general public, anti-abortion activists and the scientific community over the past few years [43]; However, the importance of fetal tissue donation to facilitate biomedical research and health care continues to this day. In 2019, the National Institutes of Health (NIH) provided $109 million in funding for 175 studies utilizing fetal tissue for a diverse array of projects encompassing infectious disease, cancer, gene regulation, neurological disorders and degeneration, retinal disease, stem cell therapy, birth defects and the mechanisms underlying normal and abnormal fetal development [44,45]. Two of the most notable projects, due to their scope and broad scientific impact, are BrainSpan [46] a highly-used public resource that depicts anatomic and molecular data for the human brain across the lifespan, and ENCODE [47,48] a large NIH-funded project that defined regulatory elements across the human genome. Ensuring that the highest quality HFT are available for research purposes also honors the intent of the tissue donation by the subject.

Strengths and limitations

This is the first study to systematically evaluate the impact of agents for the induction of fetal asystole on fetal tissue quality. Such agents were administered prior to D&E. We assessed quality using multiple measures relevant to the diagnosis of fetal anomalies and the use of HFT for research. This study has several important limitations, primarily related to the challenges of access to subjects and their donation. Our convenience sample made it impossible to fully match the delayed evacuation and immediate evacuation groups for fetal sex, gestational duration, and provider performing the termination. Ethically, because research participation cannot influence pregnancy decision-making or the termination method elected, we were unable to directly determine the effects of digoxin versus KCl, drug dose, injection site, or the interval between injection and D&E. Furthermore, our evaluation of cell viability and culturing capacity was restricted to fibroblasts and brain-derived cells (astrocytes). Other tissues and cell types were not tested, limiting the generalizability of our findings across all fetal tissues. Additionally, cell viability was only assessed in brain tissue, which does not capture potential variations in viability across other organ systems. These restrictions in methodology, while unavoidable due to sample availability and practical constraints, highlight the need for further research exploring tissue-specific differences in viability and culture success. Despite these limitations, the differences in tissue quality and viability observed between delayed evacuation and immediate evacuation HFT are so striking that they are unlikely to be due to confounding factors. Of note, this study does not directly address the impact of termination methods on diagnostic yield. However, our findings align with prior work showing that RNA degradation substantially limits transcriptomic applications [29,30,37].

Conclusion

Although our study did not specifically assess fetal anomalies alone or clinical diagnostic outcomes, it is well recognized that high‐quality fetal tissue is crucial for numerous downstream analyses, including histopathology, immunohistochemistry, and molecular profiling. To our knowledge, this is the first systematic evaluation of the effects of induced fetal demise and delayed evacuation on human fetal tissue quality. Our findings suggest that a shorter interval between the induction of fetal demise and evacuation is more likely to preserve tissue quality for both clinical and research applications. In settings where demise is induced prior to evacuation, providers may consider reducing this interval to ≤6 hours for patients who desire more reliable diagnostic evaluations. Intracardiac KCl or lidocaine injection, for instance, can be administered immediately before dilation and evacuation (D&E) or labor induction, facilitating expedited procedures [19]. Additional methods to induce fetal asystole at the time of D&E include mechanical techniques, such as umbilical cord transection, which can be completed immediately prior to uterine evacuation [21]. Ensuring that there is minimal tissue degradation can yield more promising results, enhancing the reliability of further analysis and deepening our understanding of fetal development and pathology in broader contexts.

While the study was not focused on studying fetal anomalies, a small proportion of specimens documented abnormalities (4/44, evenly distributed between delayed and immediate evacuation). Future studies investigating cases involving suspected anomalies are warranted, as tissue degradation can further hinder accurate pathological evaluation and molecular testing. Expanding sample size, comparing additional organ systems, and directly evaluating miscarriage-derived tissue will be crucial toward refining tissue handling protocols and improving outcomes for both clinical care and research.

Supporting information

S1 FigDNA quality with delayed evacuation compared to immediate evacuation.Sex-specific PCR products. Two IE specimens (C1-C2) and eight DE specimens (D1-D8) visualized on a 1% agarose E-Gel (Life Technologies). DNA integrity numbers (DIN) are shown above the gel. Specimens C2, D4, and D7-8 all had male external genitalia and amplified both X (977 base-pairs) and Y (790 and 358 base-pairs) chromosome-specific PCR products, while C1, D1-3, and D5-6 had female external genitalia and amplified only the X chromosome-specific PCR product. The DNA ladder (M) used is E-Gel 1Kb Plus DNA Ladder (Life Technologies #10488−090).(PDF)

S1 TableSummary of specimen demographics and experimental details.This table provides an overview of specimen demographics, including subject groupings (DE vs. IE), figure identifiers, DIN (DNA Integrity Number) and RIN (RNA Integrity Number) scores, DV200 scores, and the status of fibroblast culture, TUNEL, and Ki67 assays. Gestational age and post-conception days were banded using 1-week and 10-day half-open bins, respectively; Both are inclusive of the lower bound and exclusive of the upper bound.(XLSX)

S2 TableTissue quality assessment by histochemistry (TUNEL) and immunohistochemistry (Ki67).Symbols indicate the proportion of cells with stain uptake compared to control tissue: (+++) 76–100%, (++) 50–75%, (+) 25–49%, (-) 0–24%. Semi-quantitative assessments of apoptosis (TUNEL) and proliferation (Ki67) in tissues from three DE specimens evacuated 19–25 hours after digoxin injection compared to tissues from IE specimens. Semi-quantitative assessments of apoptosis (TUNEL) and proliferation (Ki67) in tissues from two IE specimens incubated at 37°C for 0, 6, and 18 hours.(PDF)

S1 DataSupplemental data.(PDF)