Advances in Fetal Surgery: Current and Future Relevance in Plastic Surgery

Abstract

Scarless healing has long been the holy grail for plastic surgery. While historically fetal surgery has tempted plastic surgeons with the allure of scarless correction of congenital abnormalities, the risks far outweighed the benefits and these interventions never materialized. Current advances in fetal surgery with minimally invasive fetoscopic surgery have made these early fetal interventions safer, leading to expanding applications. While the plastic surgeon's role is limited as of yet, this article provides a review of the history of fetal surgery and the advances that may become relevant to the future plastic surgeon.

Almost 40 years ago, the standard of care for the treatment of congenital malformations was a planned delivery at a tertiary care center where postnatal repair could be attempted. At the time of delivery, many of these anomalies would have already caused devastating, irreversible damage, while others still were fatal. Since Dr. Michael Harrison performed the first open hysterotomy fetal surgery in 1981, rapid advances in the field have allowed fetuses with malformations such as myelomeningocele (MMC), sacrococcygeal teratoma, obstructive uropathy, and congenital diaphragmatic hernia, as well as placental abnormalities such as twin-twin transfusion syndrome and chorioangioma, to be treated or ameliorated using in-utero surgery. The field relies on a multidisciplinary team including pediatric surgeons, maternal fetal medicine subspecialists, neonatologists, pediatric radiologists, sonographers, and genetic counselors. Depending on the specific anomaly, other team members can include pediatric cardiologists, neurosurgeons, nephrologists, endocrinologists, or plastic surgeons. ¹ Fetal surgery was of initial interest for plastic surgeons for cleft lip repair. While this has fallen out of favor, plastic surgeons continue to have a role in fetal treatment of amniotic band syndrome and MMC repairs and there is ongoing research into the in-utero treatment of other conditions including craniosynostosis.

Start of Fetal Intervention

The development of amniocentesis into a therapeutic tool was the first step toward fetal intervention. While amniocentesis was initially used for diagnosis only, Sir William Liley repurposed it for antenatal therapy. With the aid of fluoroscopy, he performed a transuterine fetal intraperitoneal red blood cell transfusion for erythroblastosis fetalis in 1963. ² This procedure allowed a fetus with Rh incompatibility, who was too young to survive premature delivery, to be carried to term.

Use of Technology to Further the Field

Ultrasound

The advent of the prenatal ultrasound in the 1970s made fetal imaging possible, allowed for the study of fetal pathophysiology, and ushered in advances in sampling techniques. More recently, 3D and 4D ultrasound have furthered the ability to diagnose structural abnormalities. Some applications of these advanced ultrasound techniques include calculation of fetal organ volumes, real-time reconstruction of the fetal cardiac cycle from multiple planes, ³ and calculation of ejection fraction and valvular regurgitation via Doppler flow. The greater level of detail provided by these 3D/4D ultrasounds also allows for better characterization of anomalies and is vital in fetal surgical planning ( Fig. 1 ). ⁴

Fig. 1.

Intraoperative ultrasound showing left upper arm amniotic band constriction.

By detecting hydrops, which is an indicator of imminent fetal demise, Doppler velocimetry can help determine surgical intervention timing. ⁵ Also, ultrasound is a valuable intraoperative tool which aid the surgeons by providing real-time visualization of fetal positioning and placental location, aids in selecting hysterotomy site in open surgical cases, and trocar site placement in fetoscopic cases, and allows accurate guidance for needle placement during amniocentesis and needle-based procedures.

While ultrasound technology allowed for fetal surgery's growth—aiding in diagnosis, surgical planning, and intervention technique—it does have its limitations. For example, maternal adipose tissue can cause beam attenuation, maternal pelvic bones can obscure fetal parts, and oligohydramnios can limit the image quality. ⁶

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI), specifically ultrafast MRI, which eliminates the need for fetal immobilization by capturing a sequence in 20 seconds, provides a good imaging alternative when ultrasound is insufficient. ⁷ A prenatal evaluation combining ultrasound with fetal MRI has allowed for more accurate and precise diagnosis of congenital malformations.

Animal Models

Lamb and nonhuman primate models have been used extensively to study fetal pathophysiology and intervention techniques. While multiple models demonstrated improvements in fetal animal outcomes, such as improved lung capacity in congenital diaphragmatic hernia ⁸ and superior postnatal motor skills in hydrocephalus, ⁹ there was evidence of increasing preterm delivery rates secondary to manipulation of the uterus. ¹⁰ Anesthesia protocols were developed, utilizing inhaled anesthetics (propofol for initiation followed by sevoflurane to a MAC of 1–3) as well as maternal tocolytics (preoperative oral indomethacin or nifedipine followed by postoperative magnesium sulfate and indomethacin for 48 hours) to decrease uterine activity. ¹¹ While studies in the 1980s to 1990s demonstrated that fetal surgery in humans was possible, neither survival rates nor outcomes were improved, as many fetuses had complications resulting in fetal death and surviving fetuses had outcomes similar to those treated postnatally. ^{12 13 14 15} As such, strict criteria for fetal surgery were developed, selecting only fetuses with conditions that could not be fixed postnatally and that could benefit from early fetal intervention. ¹⁶

Surgical Approach

Open Approach

Early fetal surgeries were performed by an open approach, which required both a maternal laparotomy and hysterotomy. This approach, though the most invasive, is still indicated for certain cases including MMC, congenital cystic adenomatoid malformation/congenital pulmonary airway malformations with large lung masses and associated nonimmune hydrops and cardiac failure, and large vascular sacrococcygeal teratomas with high-output heart failure. Complications of open fetal surgery include postoperative preterm uterine contractions, amniotic fluid leakage secondary to chorioamniotic membrane separation or frank membrane rupture, maternal reaction to tocolytics (pulmonary edema), abruption, uterine dehiscence, future development of placenta accreta, and need for delivery via cesarean section in the current as well as all future pregnancies. ¹¹

Closed, Minimally Invasive Approach

Closed, minimally invasive procedures avoid a hysterotomy. A needle or trocar is placed through the uterine wall with a puncture ranging from 1 to 4 mm. Most cases are performed using only one puncture site ( Fig. 2 ).

Fig. 2.

Intraoperative photograph of open fetoscopic surgery being performed with the uterus externalized. Ultrasound guidance is utilized along with the endoscope.

Some cases require an incision through the maternal abdominal wall and others are performed percutaneously. ¹¹ While there is still risk of preterm contractions, preterm premature rupture of membranes, abruption, chorioamnionitis, and fetal distress, there is less risk of uterine dehiscence and a cesarean delivery is not required unless for obstetric indications. A minimally invasive approach is used for cases of congenital diaphragmatic hernia, twin-twin transfusion syndrome, chorioangioma, vasa previa, amniotic band syndrome, lower urinary tract obstruction, critical aortic or pulmonary stenosis, and pleural effusion or other fluid-filled chest space-occupying lesions. There are two different types of minimally invasive procedure—fetoscopy and percutaneous fetal interventions guided by sonography (FIGS); however, today many procedures are “sono-endoscopic” using both ultrasound and fetoscopic imaging to perform the operation. ¹⁷

Fetal Intervention Today

After many technological and technical advancements, fetal surgery has been shown to positively affect fetal outcomes through a variety of operative techniques, ranging from outpatient, percutaneous procedures to invasive open fetal operations involving maternal laparotomy and hysterotomy. Increased knowledge about fetal pathophysiology has led to more precise intervention timing. Furthermore, the focus has shifted from the surgical correction of an anatomical malformation to the manipulation of physiologic processes to affect downstream developmental consequences of these malformations. ¹

Today, about 3% of all babies born each year have a complex birth defect. These families can seek medical care in one of several fetal surgery centers in the United States—including Baylor College of Medicine, University of California San Francisco, Children's Hospital of Philadelphia, and the Cincinnati Fetal Center. Thanks to continued animal and clinical research, with a focus on multicenter, randomized, controlled human clinical trials, in-utero surgery has advanced to a point that it is now offered for both lethal and select nonlethal malformations.

Fetal Wound Healing

The fetus has the intrinsic ability to regenerate tissue. In 1979, Rowlatt noted that fetuses at 20 weeks gestation did not form scars as normally seen in adults and, instead, healed via mesenchymal proliferation. ¹⁸ Dr. Harrison's Fetal Treatment Program at University of California San Francisco made similar observations during its initial clinical series of fetuses 18 to 28 weeks gestational age. Animal models in rats, ¹⁹ rabbits, ^{20 21 22} sheep, ^{23 24} and monkeys ^{25 26 27} have demonstrated that dermal repair is fundamentally different in fetuses compared with adults. The process by which adults undergo wound healing involves four stages—hemostasis, inflammation, proliferation, and remodeling. ²⁸ The wound healing process is performed via an intricate coordination of physical, biological, and chemical signals that culminates in a fibrotic scar formation, where the damaged tissue is never fully stored to its preinjury state.

Wound Healing in Adults

Initially after tissue injury in adults, an inflammatory and hemostatic phase ensues. Hemostasis is achieved via two main mechanisms—vasoconstriction of the small vessels within the wound and activation of the coagulation cascade via tissue factor and intrinsic pathways leading to platelet aggregation and degranulation. As a result of this initial response to tissue injury, a fibrin matrix forms that provides a scaffold for later stages in the healing process. Furthermore, platelet degranulation results in the release of growth factors, chemotactic agents, vasoactive mediators, and other cytokines (platelet derived growth factor, transforming growth factor α and β, and epidermal growth factor) that are essential to wound healing. The released growth factors and cytokines recruit inflammatory cells to the wound site—neutrophils are the first inflammatory cells to arrive, followed by macrophages and lymphocytes. The macrophage, the essential inflammatory cell of wound healing, will secrete enzymes and cytokines which promote transition into the second stage of wound healing.

The fibroproliferative phase, which results in epithelialization, angiogenesis, granulation tissue formation, and collagen deposition, starts three days after tissue injury and lasts about two weeks. During this phase, fibroblasts migrate and deposit extracellular matrix proteins (hyaluronan, fibronectin, proteoglycans, and type 1 and 3 procollagen), replacing the temporary fibrin meshwork with collagens. In the final stage of wound healing, wound maturation and remodeling occurs, which results in a wound that will have an eventual strength of ∼80%.

Fetal Wound Healing

The exact mechanism intrinsic to fetal skin that allows scarless healing is unknown. It was theorized that differences in the fetal wound environment and/or fetal cells themselves are responsible for the ability of fetuses to heal without any macroscopic evidence of dermal injury. Fetal wounds differ from those in adults with regard to inflammatory response, extracelllar matrix makeup, growth factor and cytokine expression, signaling cascades, and gene expression profiles.

Histologically, fetal skin is initially a single layer of ectodermal cells starting at day 7 or 8 of gestation. It then becomes stratified from 9 to 14 weeks' gestational age, undergoes follicular keratinization from 14 to 24 weeks, and interfollicular keratinization after 24 weeks. ^{29 30} The ability for scarless wound healing is present prior to 24 weeks gestational age. ^{24 31 32} Lorenz et al used athymic mice with transplanted human skin to determine this transitional gestational age after which point healing occurred in a postnatal manner resulting in excess collagen in the extracellular matrix (ECM), loss of dermal appendages, and flattened epidermis. ³¹ The size of the wound also determines its scarring outcome with larger wounds requiring earlier repair to avoid scar formation. ³³

While fetuses reside in a sterile and hypoxic intrauterine environment with amniotic fluid that is rich in growth factors, fibronectin, and hyaluronic acid ^{34 35} and a pO ₂ averaging 20 mm Hg ³⁶ (three to four times less than that of adults), studies by Longaker and Lorenz demonstrated that factors intrinsic to fetal cells were more significant to wound healing than extrinsic environmental factors. ³⁷ As in adults, fibroblasts deposit ECM; however, fetal fibroblasts synthesize a greater total amount of collagen as well as a higher ratio of type III to type I collagen (30–60% type III collagen in fetuses, 10–20% in adults). ^{38 39 40} As gestational age increases, the amount of collagen crosslinking also increases, causing increased matrix rigidity and its associated scar formation. ^{24 41 42} Furthermore, the fibroblast cell surface receptors that collagen binds to, discoid domain receptors (DDRs), differ based on gestational age with DDR-1 expressed largely in early gestation. ⁴³ This difference in tyrosine kinase receptor expression may contribute to fetal wound healing. ⁴⁴

Scarless fetal wounds also have few if any myofibroblasts. This lack of contractile forces by the myofibroblasts as well as the collagen type profile allows for collagen fibrils to be oriented in a fine, reticular, or basket weave, rather than in dense parallel bundles arranged perpendicular to the wound surface. ⁴⁵ Early gestation fibroblasts are also able to quickly migrate to a wound site and multiply at the same time they produce collagen, thus avoiding a delay in collagen deposition. ⁴⁶ Fetal fibroblasts also have greater expression of hyaluronic acid (HA) synthases, an increased number of HA receptors, as well as a muted response to cytokines. ^{47 48} Studies on fetal lambs have shown that HA is deposited early on and in both a rapid and sustained fashion in fetal wounds. Longaker showed that the level of HA as well as HA-stimulating activity (HASA) significantly decrease during the transitional period in fetal gestation (120 days in lamb model; 24 weeks in humans). ⁴⁹ Increased HA allows increased cell density, increased collagen III deposition, and accelerated wound repair. As HA content decreases later in gestation and into the postnatal period, there is an appearance of other extracellular proteoglycans (decorin and heparan sulfate) along with the onset of scarring. ⁴⁵ Certain glycosaminoglycans and proteoglycans such as these are present largely in adult wounds, whereas others such as chondroitin sulfate and fibromodulin are present in fetal wounds and absent in adult wounds and may play a role in scarless regeneration. Fibromodulin specifically binds to and inactivates transforming growth factor (TGF)-β. ⁵⁰

Early gestational fetal wounds also have different ECM adhesion proteins and cell surface receptors, with fibronectin and tenascin being rapidly upgraded and deposited in large quantities. ^{51 52} This results in early cell attachment and migration allowing for an organized wound matrix to be formed. Epidermal integrin receptors on fetal keratinocytes near wound edges are also upregulated, leading to increased keratinocyte proliferation, migration, and a hastened epidermal repair. ⁴⁵ Lastly, the relative activity of matrix metalloproteinases (MMPs) and tissue-derived inhibitors of metalloproteinases (TIMPs) play a role as intrinsic fetal proteins within the ECM. MMPs degrade a wide spectrum of ECM proteins and TIMPs counteract their proteolytic activity. As the expression of these proteinases differs, there is an effect on ECM remodeling and wound healing and in fetal wounds matrix turnover occurs both rapidly and efficiently. ^{53 54 55}

Another set of characteristics intrinsic to fetal cells involves the inflammatory response. Prior to the transitional period at ∼24 weeks gestation, fetuses are able to rapidly re-epithelialize their skin with a minimal acute inflammatory response as there is a very limited presence of neutrophils. The lack of polymorphonuclear leukocytes in fetal wounds varies significantly from the inflammatory response in adults. As stated previously, in adults, platelets release TFG-β1 and platelet-derived growth factor which, along with TNF-α and IL-1, work to stimulate neutrophil migration and cause neutrophil upregulation of adhesion molecule expression. Fewer of these growth factors are present in fetal tissues and, therefore, there is decreased neutrophil migration to the fetal wound bed and a resultant diminished inflammatory response. ⁵⁶ Moreover, there is a decrease in macrophages as TGF-β1 levels are low in fetuses, ⁵⁰ resulting in decreased conversion of monocytes to macrophages. There is also a lower volume of less mature mast cells in fetal wounds which leads to a decrease in neutrophil recruitment, a decrease in fibroblast transformation to myofibroblasts, and a decrease in MMP release. ⁵⁷ Other key differences in fetal wound healing compared with adults include higher levels of mesenchymal stem cells with accompanying E-cadherin-positive cells in fetuses. ⁵⁸ While many of these factors likely play a role in scarless fetal wound healing, there still remains a relatively poor understanding of the precise mechanisms and their roles. However, the phenomenon has many potential applications, especially in the field of plastic surgery.

History of Plastic Surgery in Fetal Intervention

Cleft Lip and Palate

Malformations of the face such as cleft lip and palate have been studied as candidates for fetal intervention because in-utero repair can allow anatomic reconstruction that focuses not only on function but also aesthetics. Since wounds are repaired via regeneration prior to 24 weeks gestational age, this provides a window of opportunity for the fetus to undergo repair with reduced lip scarring and less midfacial growth retardation. ⁵⁹ Several studies have looked at wound healing in fetal bone and mucosa and found that such wounds heal in utero without formation of abnormal tissue (callus). The fetal lamb model has been preferred due to its longer gestational period, resistance to preterm labor, and ease of surgical manipulation. Lambs do not intrinsically form defects in response to phenytoin or cyclophosphamide teratogen exposure, so defects must be created surgically, which allows for the creation of standardized defects in size and severity between study subjects. Studies in this animal model have demonstrated that repairs heal without visible scar formation or inhibition of maxillary or mandibular growth postoperatively. ^{60 61 62 63 64 65 66} Unfortunately, the ideal gestational age for optimal wound healing coincides with a gelatin-like consistency of skin which makes suture placement very difficult. ⁶⁷ Sullivan proposed using a special clip that joins together the cleft lip margins as an alternative to suturing. ⁶⁸ A subsequent study compared an endoscopic in-utero approach versus the standard technique with the use of clips. Results showed an advantage in both time of procedure and inflammatory response with clip usage. ⁶⁹

Another concern with fetal cleft repairs is the development of nasal malformations after cleft correction. ^{66 67 70} A subsequent study by Levine showed that nose shape can be modified using hyperosmolar sponge fragments without risk of scar formation. ⁷¹ To date, only one case of unilateral cleft lip repair has been performed on a human fetus. The case was on a 19-week-old fetus and an open technique was used with a rotation advancement lip repair. Unfortunately, while there was no postoperative scar noted, the child was born prematurely, spent two months in the neonatal intensive care unit and ultimately did not survive. ⁷² While the prospect of fetal intervention for cleft lip and/or palate repair is promising, it is a non–life-threatening anomaly with successful postnatal care timelines and, therefore, any intervention must be held to the highest standard. Further human fetal repairs have not yet been attempted and experimental nonhuman models are still being examined.

Craniosynostosis

The treatment of craniosynostoses was another area of plastic surgery explored for fetal intervention. A trial on fetal lambs using Gore-tex mesh implants after resection of the obliterated suture was performed with results showing that all four treated lambs showed no evidence of bone regeneration and retained patent craniectomy sites. ^{73 74} In-utero correction also decreased the severity of deformities associated with the disorder including orbital position, skull length, and frontal bone shape. However, early diagnosis would be necessary for fetal intervention which is not yet performed reliably. Moreover, endoscopic repair of single suture craniosynostosis combined with helmeting has shown great results when performed early postnatally. Similar to fetal interventions for cleft lip and palate, the benefits of fetal interventions for craniosynostosis may not outweigh the risks for this treatment. Nevertheless, as fetoscopic surgery continues to develop and increase in safety, these interventions may some day find their place.

Plastic Surgery's Current Role in Fetal Intervention

Myelomeningocele Repair

In-utero treatment of MMC was the first fetal surgery to be performed for a nonlethal, yet highly morbid condition. MMC is the most common congenital birth defect of the nervous system with an incidence of 1 in 2,900 births in the United States. ⁷⁵ The mechanism of MMC injury follows a two-hit hypothesis. The first hit is a defect in neurulation resulting in incomplete closure of the neural tube with subsequent exposure of neural tissue, cerebrospinal fluid leakage, and hindbrain herniation. The second hit involves the exposure of neural tissue to the hypotonic amniotic fluid later in gestation as well as direct trauma to the exposed neural tissue by the uterine wall as the fetus moves. ^{76 77} This second hit is one that fetal intervention could theoretically avoid by covering the exposed fetal spinal canal and preventing continued spinal cord damage. Fetal lamb models have shown significant improvements in neurological outcomes in those treated in utero—results included improved motor, bladder, and bowel function and a reversal of Arnold–Chiari malformation. ⁷⁸ The MOMS trial in 2011, a randomized controlled study of fetal (open) versus postnatal MMC repair revealed a decreased need for ventriculoperitoneal shunting as well as improved functional and neurological outcomes, thus demonstrating the efficacy of prenatal surgery. ⁷⁹

Many early fetoscopic and open-repair cases used a three-layer microneurosurgical repair where the neural placode is first dissected, untethered, and retubularized, the dura is then reapproximated, and finally wound closure is achieved by either undermining of the skin or by using an amnion patch if skin alone is inadequate. Unfortunately, half of the surviving patients required wound revisions within the first seven postnatal days. ⁸⁰

Early attempts at endoscopic patch coverage of MMCs using a maternal split thickness skin graft were met with little success. ^{80 81} Some animal and human studies have also explored proximally and distally based autologous multilayer latissimus dorsi flaps to cover MMC defects. ⁸² These fetal flap surgeries, however, require extensive fetal exposure and have not been performed in humans. More recent fetoscopic meningomyelocele repairs have involved both single-layer and multiple-patch closures and have been much more successful in terms of both maternal and neonatal outcomes ^{83 84} than those initially reported by Bruner et al and Farmer et al, likely due to advances in equipment, surgical approach, and anesthesia techniques. In some cases Belfort et al ⁸³ use relaxing incisions in the fetal flanks to create a flap to allow for a watertight closure of the skin over large myeloschisis lesions where large skin deficits exist. Such relaxing incisions may occasionally require plastic surgery repair after birth when the fascia has been breached, but usually heal very well by secondary intention. ⁸⁵

Open hysterotomy fetal surgery for neural tube defect now employs a technique developed at CHOP using myofascial flaps from the paraspinal muscles sutured over the defect, with skin closure over the flaps. ⁸⁶

Current involvement of plastic surgeons is for postnatal wound revisions when necessary.

Amniotic Band Syndrome Dissection

Amniotic band syndrome, where fetal parts become entrapped in strands of amniotic tissue, has an incidence of 1 in 3,000 to 15,000 live births. While the exact pathogenesis of such bands is unknown, it is hypothesized that partial rupture of the amniotic sac early in gestation leads to the formation of fibrous bands that float in the amniotic fluid and can encircle parts of the fetus. When fetal limbs are constricted by these amniotic bands, vascular perfusion is compromised, leading to limb deformity or autoamputation. Fetal death can also occur secondary to umbilical cord strangulation. ⁸⁷ Umbilical cord strangulation is reported in up to 10% of amniotic band syndrome cases. ⁸⁸ In-utero release of amniotic bands using a minimally invasive, fetoscopic approaches have now been performed with success where lasers or scissors are used to cut the bands, restore blood flow to the distal appendage, and allow for preservation of limb function ( Fig. 3 ). ⁸⁹

Fig. 3.

( a ) (left) Fetoscopic division of left upper arm amniotic band. ( b ) (right) Residual edema of left upper arm after division of constrictive amniotic band.

Fetal intervention allows for continued development of the fetal part as well as improved function and aesthetic appearance. ⁹⁰ Risks to such intervention include preterm premature rupture of membranes and spontaneous preterm birth. A report by Javadian et al reports that 50% of cases result in preserved limb function. ⁹¹ The latest advance in amniotic band fetal surgery involves the use of CO ₂ insufflation fetoscopic surgery to allow for resection of the bands when access to the limbs or bands is not possible using a single port in amniotic fluid approach. ⁹² These surgeries are currently being performed at our institution with a team of fetal and pediatric surgeons. The role for plastic hand surgeon involvement is continuing to be explored in this setting.

Conclusion

Since its conception over 50 years ago, the field of fetal surgery has had dramatic expansions and changes. Technology and technique innovations have allowed for the increasing in-utero treatment of nonlethal malformations. With increasing research, more prenatally diagnosed malformations are becoming candidates for fetal intervention and eligibility criteria based on conscious risk-benefit analyses are continuously evolving. As the field continues to expand through innovation, the role of plastic surgery and the timing of intervention for congenital abnormalities will continue to evolve.