Management dilemma in Thoracoamniotic Shunt Migrations

Abstract

Shunt migration is a rare but significant complication of thoracoamniotic shunting, an intervention widely used for fetal pleural effusion. We describe a case of a term infant noted antenatally to have fetal hydrothorax that was managed with thoracoamniotic shunting but complicated by shunt migration. We also present the current literature on risk factors, complications and management of intrathoracic shunt migration. The majority of shunt migration cases are managed conservatively with no untoward postnatal sequelae, but surgical removal of the migrated shunt has been used for associated clinical complications, if visceral damage is suspected or if postnatal thoracic surgery is indicated for other reasons. We advocate an approach of conservative management for asymptomatic infants, where possible, to avoid unnecessary surgical and anaesthetic risks to very young, often already compromised children. However, further studies are still required to determine optimal management after shunt migration has occurred to ensure the best outcome.

Background

Fetal pleural effusion or hydrothorax is a non-specific fluid accumulation in the fetal pleural space. The incidence is estimated to be 1 in 10 000–15 000 pregnancies with a slight male preponderance.^{1 2} Primary fetal hydrothorax involves lymphatic fluid leakage into the pleural cavity from abnormal lymphatic development, while secondary causes include congenital lung lesions or cardiac anomalies. Such effusions may restrict lung growth and obstruct venous return resulting in hydrops. Fetal hydrothorax with hydrops is associated with a mortality rate of 75%.² Thoracoamniotic shunting (TAS) is an established procedure whereby a drainage tube is introduced into the effusion in utero through the mother’s abdominal wall. The 2006 NICE guidelines recommend consideration of shunting in cases of hydrops and at gestational ages during which lung development is occurring.³

Shunt migration is a known complication of TAS. While extra thoracic migration is associated with less clinical risk and easier retrieval at birth, intrathoracic migration has been associated with recurrent pleural effusions, tension pneumothorax and even fetal demise.^4–7 A recent single-centre retrospective study found an 8.5% rate of shunt migration.⁷ Despite its clinical significance, current literature is still lacking and there have been fewer than 50 studies since 1986 reporting thoracic shunt migration and its management. In the following report we describe a case of a hydrops-related fetal pleural effusion treated with TAS which was complicated by shunt migration requiring repeated insertions. We also present a literature review on the reported cases and management of shunt migration following TAS.

Case presentation

A primigravida in her 30s received antenatal care from 8.0 weeks gestation. Her routine antenatal blood test results were unremarkable. Non-invasive prenatal testing at 12.4 weeks showed low probability for trisomies 13, 18 and 21. Routine fetal survey at 19.4 weeks showed a large left pleural effusion with rightward displacement of the heart, but no other anomalies. Her antenatal care was subsequently transferred to the Fetal Medicine Department at a tertiary maternity hospital. Repeat ultrasound scan performed at 19.9 weeks confirmed a large left fetal pleural effusion with displacement of a structurally normal heart without any other signs of hydrops. Amniocentesis performed at 20.1 weeks revealed a male fetus with no aneuploidy. The pleural aspirate from a contemporaneous thoracocentesis was turbid and had a pH of 8, with red blood cell count of 9.45×10¹²/L, white blood cell count of 4343×10⁹/L, total protein of 18.2 g/L and glucose value of 4.8 mmol/L. Polymerase chain reaction (PCR) testing for congenital viral infections were all negative. Serology testing ruled out immune hydrops. Kleihauer-Betke test was negative. Doppler evaluation of the peak systolic velocity in the middle cerebral artery excluded fetal anaemia. A presumptive diagnosis of primary fetal hydrothorax was made.

Follow-up ultrasound scan at 22.7 weeks showed reaccumulation of the left pleural effusion and progression to hydrops fetalis with scalp edema and ascites but no pericardial effusion. In view of the worsening fetal condition and parental wishes to continue the pregnancy a thoracoamniotic shunt was inserted at 23 weeks. A Harrison fetal bladder stent (Cook Medical), double pigtail 5Fr stent was used. Two shunts were inserted under ultrasound guidance in the first sitting. The first shunt was dislodged into the pleural cavity during the procedure, and so a second shunt was inserted which was deployed into the correct position. However, a follow-up scan a week later showed worsening hydrops with the development of bilateral pleural effusions, persistent scalp edema and ascites. A further shunt was then placed into the left pleural effusion. Subsequent scans showed gradual reduction in both pleural effusions with total resolution of the hydrops by 35.7 weeks. The displaced shunt was consistently seen in the left fetal chest and the other two shunts remained in situ.

At 39.0 weeks’ gestation, she had a vacuum-assisted delivery of an apparently healthy male infant weighing 2955 g with Apgar scores of 9 at both 1 and 5 minutes of life. He had no physical signs of hydrops and did not require any respiratory support after birth. Two shunts were found at delivery, one embedded in the placenta and the other floating in the amniotic fluid, possibly dislodged by the delivery process itself. The baby had a functionally and structurally normal heart.

Investigations

Postnatal localisation of the retained shunt was done initially by chest X-ray (CXR) (figure 1A) and later by ultrasound, which showed a small calibre tube within the pleural space of the lateral left upper chest with a trace of pericardial fluid (figure 1B).

Figure 1.

(A) CXR post-delivery showing left-sided intrathoracic shunt projected in the left upper to midchest. (B) Targeted ultrasound scan of the left chest, showing small calibre tubing estimated to measure 2 mm in diameter. (C) Repeat CXR done during infancy showing intrapleural shunt in left upper midthoracic region (anterior–posterior view). (D) Repeat CXR done during infancy showing intrapleural shunt in left upper midthoracic region (lateral view). CXR, chest X-ray.

Treatment

The term neonate remained clinically stable throughout his inpatient stay. Pediatric surgeons were consulted for opinion on removal of the intrapleural shunt. After discussion with parents on risks, benefits and available options, decision was made for conservative management with deferral of surgical removal till after toddler years, when general anesthetic risks are lower, should the patient continue to remain asymptomatic and well. Repeat CXR done at his infancy follow-up showed the intrapleural shunt to be in the same position (figure 1C and D). The child remained asymptomatic at his subsequent follow-ups and eventually underwent successful and uneventful elective thoracoscopic removal of the retained shunt in his early childhood years due to parental preferences

Outcome and follow-up

The child tolerated the procedure well with no intraoperative or immediate postoperative complications noted during his 1 month postoperative review.

Discussion

Our literature review identified 16 relevant studies (figure 2) which are summarized in online supplemental table S1. There was a total of 47 fetuses among 16 included studies, all with antenatally diagnosed intrathoracic shunt migration. Most studies described the use of double pigtail catheters as shunts,^{1 7–13} although some used a double-basket catheter.^{4 14 15} Reported complications included recurrence of the pleural effusion in one fetus,⁸ pleural adhesions in two fetuses,^{1 8} tension pneumothorax in two fetuses due to unintended communication with the pleural cavity,^{4 5} as well as suspected rib fracture inferior to the site of insertion which healed in utero in one fetus.¹⁶ Shunt migration was found to occur to the right lung horizontal fissure,¹⁰ lung cyst,¹⁷ lung hilum,⁶ pleural space,^{1 8} subcutaneous skin layer,¹⁶ as well as along the course of intrathoracic aorta.¹⁰ More than half (28/47, 60%) of the patients had the shunt left in situ.1 6–9 11–13 18 When surgery was undertaken, 12 infants had shunt removal within 7 days of birth.1 4 5 10 14 16 17 For three cases, surgical retrieval was done later, one because of infant instability, another electively and one being indicated for developing concern over the shunt encircling a major vessel.^{7 10} Most infants with displaced shunts had an uneventful postnatal stay with resolution of pleural effusion. Five neonatal deaths were reported; one due to shunt compression of the lung hila,⁶ another due to pulmonary hypoplasia with the background of fetal hydrops¹ and three deaths were due to dislodged/malfunctioning shunts which were not replaced. Nine studies with 21 fetuses reported long-term follow-up, with the oldest being 22 years of age.1 4 5 7 11–13 17 18

Figure 2.

Flow diagram of included studies (illustrated by APPT and by BT).

Pathophysiology of fetal hydrothorax

A primary fetal pleural effusion is due to leakage of lymphatic fluid into the pleural cavity from abnormal lymphatic development, with most cases being diagnosed in the second to third trimester. Secondary fetal pleural effusions, which are usually associated with either immune or non-immune fetal hydrops, have many etiologies including congenital infections, congenital lung lesions such as cystic adenomatoid malformations, congenital diaphragmatic hernia or cardiac structural defects and rhythm anomalies. Some remain unexplained as in our case, but the number of these is reducing due to ever-advancing scan technologies aiding diagnosis. Such effusions can restrict fetal lung growth leading to pulmonary hypoplasia, compress the heart or obstruct the venous return to the heart resulting in hydrops, or compress the esophagus causing polyhydramnios.

Intrauterine management of fetal hydrothorax

Prenatal interventions are aimed at draining the effusion to allow normal lung development and prevent hydrops due to its cardiac effects. TAS is an established procedure whereby a metal cannula on a trochar is introduced through the fetal chest wall into the effusion in the uterine cavity through the mother’s abdominal wall. The trochar and cannula are then removed, and the drainage catheter (or shunt) is left in the chest wall with one end in the pleural cavity of the fetus and the other in the amniotic cavity. After delivery, the chest drain is immediately clamped and removed to prevent the development of a pneumothorax.

Complications of TAS

Major complications of TAS placement include preterm births, which may occur up to 80% of the time, and preterm premature rupture of membranes in up to 15% of the cases.^{19 20} These complications may be related to the procedure itself, the underlying comorbidities or both. Fetal complications following shunt insertion include constriction band of the limbs,²¹ rib fractures,^{16 22} traumatic hemothorax,¹² hypoproteinaemia²³ and even fetal deaths due to placental abruption and traumatic cord accidents.^{20 24} Reported maternal complications include scars,²⁵ uteroperitoneal leakage, development of transient maternal ascites and oligohydramnios.²⁶ Shunt-related complications include shunt dislodgement, retention, or blockage.²⁷

Shunt migration has been reported to occur between 8.5% and 20% of cases.^{7 9 13} Factors associated with migration may be shunt-related or fetal-related. Migration has been associated with thinner and softer catheters with smaller diameters.¹⁵ Our review suggests that shunt migration can occur with either double pigtail shunts with coiled ends or double basket catheters which are short and straight. Both misplacement of the device at initial insertion^{10 28} and multiple insertions,¹⁴ as in our case, have been associated with higher migration rates. However, the consideration of multiple insertions needs to be balanced with the benefits of minimising pulmonary hypoplasia and hydrops risk.⁸ Fetal factors include size, gestational age and fetal respiratory movements, with younger and smaller fetuses having more respiratory movements.⁴ Our fetus was only at 23 weeks’ gestation at initial insertion.

Our case also highlights an additional use of ultrasound to complement radiological imaging for localisation in the event of migrated shunt. It is useful for monitoring progress and prognostication of fetal primary hydrothorax antenatally, as shown in our case, with effusion-to-chest ratios of 0.4 or greater signifying potentially hydropic fetuses.²⁹ Ultrasound has also been found to be useful in retrieving and replacing dislodged shunts.³⁰ It remains the modality of choice in young patients due to its capabilities for dynamic and high spatial resolution imaging, with added advantages of being radiation-free and portable.

Management of dislodged or retained TAS

Different strategies have been proposed for the management of a dislodged TAS. Our review showed that conservative management was favoured by a larger proportion of studies, with many well-thrived infants at long-term follow-ups. However, opinions on management do vary with one study promoting conservative management due to inert material used for shunts,¹² while another recommending earlier surgical removal to minimise the risk of fibrous and inflammatory response to the foreign body, potentially making removal more challenging.¹⁰ Another interesting study also highlighted the positive correlation between the degrees of intrathoracic inflammation in cases of fetal pleural effusion with poorer neonatal outcomes.³¹ Surgical management is further advocated if there are suspicions of visceral injury or unintended communications from malpositioned shunts, or when a thoracic surgical procedure is nonetheless required postnatally.5 10 14 32 A recent large-scale study on TAS dislodgement supports conservative management if postnatal imaging excludes mediastinal involvement.⁷ Currently, there are no standardized guidelines for the management of a migrated shunt. Opting for conservative management may raise concerns regarding the potential detrimental impact of a retained shunt, secondary to triggered sustained inflammatio, which may hence influence the long-term prognosis of the child. Further research is needed to investigate the properties of thoracoamniotic shunts and degree of associated inflammatory response. There is also a lack of long-term follow-up data from cases after the surgical removal of migrated shunt. Further studies with careful patient selection and close follow-up for complications are required to determine the most appropriate management. Weighing the risks and benefits of putting a young infant through surgery under general anesthesia, we recommend holding off surgical removal of the retained shunt especially if the infant remains clinically well.

Learning points.

• Thoracoamniotic shunt insertion is the intervention of choice for the management of fetal hydrothorax especially when it is isolated and associated with hydrops fetalis.

• Thoracoamniotic shunt migration can result in significant mortality and morbidity.

• In an otherwise asymptomatic child with retained thoracoamniotic shunt, multidisciplinary comanagement is essential to decide on optimal timing of surgery when the child reaches an age at which operative and anesthetic risks may be better tolerated.

Ethics statements

Patient consent for publication

Consent obtained from parent(s)/guardian(s).