ERCP in Pregnancy

Abstract

Background: Although rare, pancreaticobiliary disease during pregnancy can pose a serious risk to both the mother and fetus. Data regarding the relative safety of endoscopic retrograde cholangiopancreatography (ERCP) during pregnancy are sparse. Methods: We performed a retrospective review of 17 ERCP procedures performed at a single tertiary care referral center between January 2005 and April 2009. Records were reviewed for ERCP indication, endoscopic interventions, use and extent of fluoroscopy, postprocedure complications, and pregnancy outcomes including Apgar scores. Results: All procedures were performed without any maternal adverse events immediately or on follow-up. There were no signs of fetal distress during any of these cases, and there were no fetal complications noted upon delivery or at 30-day follow-up per chart review. Conclusion: Therapeutic ERCP during pregnancy appears to be safe when performed in experienced hands and with judicious use of fluoroscopy.

Pancreaticobiliary disease in pregnancy is common and can lead to significant maternal and fetal morbidity. Gallstone disease is estimated to complicate approximately 3–12% of pregnancies,^1,2 and 1 out of every 1,000–1,200 pregnancies is complicated by symptomatic biliary disease.^2,3 Choledocholithiasis is a serious condition leading to potentially life-threatening consequences such as cholangitis or gallstone pancreatitis. Surgical therapy has traditionally been considered the treatment of choice for gallstone disease but has been associated with an increased risk of fetal compromise.⁴

Endoscopic therapy with endoscopic retrograde cholangiopancreatography (ERCP) has been suggested as an effective alternative therapeutic option for managing pancreaticobiliary diseases during pregnancy. Data are limited, however, regarding the safety of ERCP during pregnancy.

We report our experience with 17 ERCP procedures performed in 10 pregnant patients from a single tertiary care referral facility.

Patients and Methods

Seventeen ERCP procedures (performed in 10 pregnant patients between January 2005 and April 2009) were identified retrospectively. ERCP was performed utilizing Olympus TJF-160 and TJF-160VF videoendoscopes.

Indications included cholangitis (n=1), gallstone pancreatitis (n=4), obstructive jaundice (n=5), stent exchange or removal (n=5), and cystic dilation of the biliary tree with elevated transaminases (n=2). Clinical suspicion of pancreaticobiliary disease requiring endoscopic intervention was confirmed with either preprocedure ultrasonography or magnetic resonance cholangiopancreatography (MRCP) in all patients.

All procedures were performed by experienced endoscopists under propofol sedation with standard cardiopulmonary monitoring with the assistance of an anesthesiologist or certified nurse anesthetist. Prophylactic antibiotics were administered to all patients in anticipation of the possible need for sphincterotomy or other therapeutic interventions. Maternal-fetal monitoring was performed throughout all procdures, with documentation of fetal heart tones prior to sedation and immediately upon completion of the procedure. Fetal radiation exposure was not assessed during any of the procedures, and procedure times were not readily available for the majority of procedures.

Patients were assessed for maternal complications as well as fetal heart tones immediately postprocedure and were monitored postprocedure in the hospital on the obstetrics service. Post-ERCP complications were evaluated using consensus definitions⁵ and defined as any adverse events such as fever, pancreatitis, or bleeding, in addition to signs of fetal compromise. Fetal complications were assessed at delivery and at 30 days follow-up postdelivery.

Results

The mean patient age was 23.5 years (range, 18–30 years), and the mean duration of pregnancy was 20 weeks (range, 8–34 weeks). Therapeutic ERCP was performed in all patients (Table 1). Biliary stones or sludge was removed in 9 cases, and sphincterotomy was performed in 10 cases.

Table 1.

Endoscopic Retrograde Cholangiopancreatography (ERCP) and Associated Interventions

Patient	Indication	ERCP	Trimester	Fluoroscopy	Intervention
1*	Cholangitis	1	1	No	Biliary stent placement
	Stent exchange	2	2	No	Stent exchange
	Stent exchange	3	2	No	Stent exchange
	Stent exchange	4	3	No	Stent exchange
2	Gallstone pancreatitis		3	No	Sphincterotomy, balloon-stone extraction
3*	Obstructive jaundice	1	2	Limited	Sphincterotomy, balloon-stone extraction, biliary stent
	Stent exchange	2	2	No	Stent exchange
	Stent exchange	3	3	No	Stent exchange
4	Gallstone pancreatitis		3	Limited	Sphincterotomy, balloon-stone extraction
5	Gallstone pancreatitis		2	No	Sphincterotomy, balloon-stone extraction
6	Obstructive jaundice		1	No	Sphincterotomy, balloon-stone extraction
7	Obstructive jaundice		1	Limited	Sphincterotomy, balloon-stone extraction
8†	Biliary dilation	1	1	No	Sphincterotomy, pancreatic duct stent
	Biliary dilation	2	1	Limited
9	Obstructive jaundice	1	3	No	Sphincterotomy, balloon-stone extraction, biliary stent
	Obstructive jaundice	2	3	Limited	Stent removal, sphincterotomy, balloon-stone extraction
10	Gallstone pancreatitis		1	Limited	Sphincterotomy, balloon-stone extraction

^*Stone definitively removed postpartum.

^†ERCP revealed a choledochal cyst.

Limited use of fluoroscopy was performed in 6 cases, with a mean exposure time of 8 seconds. Lead shielding to the pelvis was provided in all cases involving fluoroscopy in an attempt to reduce radiation exposure to the fetus. Wire-guided cannulation was utilized in all cases. Aspiration of bile was used to confirm selective bile duct cannulation in cases where fluoroscopy was not used. In patients with large biliary stones (>10 mm), or cases in which there was some doubt regarding complete stone clearance, a plastic biliary stent was placed to ensure bile drainage while preventing stone impaction.

Patient #1 (on Table 1), who initially presented in the first trimester with cholangitis, was noted to have a large stone in the common bile duct with a dilated biliary system observed on MRCP. ERCP was subsequently performed with the placement of a 10 French plastic biliary stent to allow for bile drainage. Stone extraction was not attempted at that time, given the possibility of stone impaction due to the size of the stone. In addition, lithotripsy was not attempted given the possibility of inducing pancreatitis with excessive duct manipulation and trauma as well as the remote possibility of impacting the stone at the ampullary orifice if lithotripsy was unsuccessful. Three subsequent ERCP procedures were performed in this patient with stent exchange. All of these procedures were performed without fluoroscopy and utilized wire-guided cannulation and bile aspiration techniques for confirming biliary cannulation. The patient returned in the postpartum period for definitive stone extraction.

Similarly, another patient presenting with obstructive jaundice, Patient #3, underwent ERCP with a similar therapeutic plan in which a 10 French plastic biliary stent was placed for biliary decompression and prevention of gallstone pancreatitis. A repeat ERCP was performed postpartum with definitive stone extraction.

Additionally, a patient presenting with obstructive jaundice in the setting of cholelithiasis, Patient #9, underwent ERCP with sphincterotomy and balloon extraction of a small stone. Fluoroscopy was not utilized. A 10 French plastic biliary stent was placed, given the presence of cholelithiasis. Despite the stent placement, the patient continued to have protracted jaundice, and a repeat ERCP performed with spot fluoroscopy revealed a stone in the intrahepatic biliary system. The previous sphincterotomy site was extended, and the stone was extracted via balloon, with subsequent normalization of bilirubin levels. The patient delivered several weeks later, and a cholecystectomy was performed postpartum.

Another patient (Patient #8) underwent ERCP to evaluate abdominal pain associated with transaminase elevation and cystic dilation of the biliary system noted on MRCP. On initial ERCP, biliary cannulation was difficult and required temporary pancreatic duct stenting and subsequent access needle-knife sphincterotomy. Despite these attempts, biliary cannulation was unsuccessful. A repeat ERCP was performed with minimal contrast injection and spot fluoroscopy revealing an anomalous union of the pancreatic duct and common bile duct with stenosis of the distal bile duct and proximal cystic dilation suggesting a type IA choledochal cyst. As we were unable to cannulate the bile duct proximal to this stenosis, a biliary stent was not placed. The patient’s pregnancy was uneventful thereafter, without any signs of fetal distress or compromise. She later opted to forego pregnancy and electively aborted the fetus. The patient was then lost to follow-up.

No maternal adverse events were noted immediately after any of these procedures or on follow-up. In addition, there were no signs of fetal distress during any of these cases. Apgar scores noted upon delivery were greater than 8 at 1 and 5 minutes. Infants appeared to be healthy upon chart review of records from 30-day postpartum follow-up. Long-term follow-up, beyond 30 days, was not available in any of these cases.

Discussion

Symptomatic pancreaticobiliary disease during pregnancy is relatively common and can pose significant risk of detriment to both mother and fetus. The most common of these diseases include symptomatic choledocholithiasis and gallstone pancreatitis.¹ Therapeutic ERCP in this setting becomes a vital component of our armamentarium for treating these disorders. The first report of ERCP during pregnancy was published in 1990 by Baillie and colleagues.⁶ Since then, several case reports and series have been reported in the literature documenting the safety of therapeutic ERCP in pregnancy.^2,3,7-12 One such study focused on the use of bile aspiration techniques and wire-guided cannulation as an alternative to fluoroscopy.²

It has been suggested by some that ERCP is likely best performed during the second trimester, though the procedure appears reasonably safe to perform throughout the entire period of pregnancy.¹³ This suggestion is based upon historical data obtained from the surgical literature. The consensus among surgeons is that the second trimester is the safest period to perform surgery during pregnancy, as there is a high rate of fetal loss in association with first trimester surgical intervention.¹⁴ This risk has not been clearly shown with endoscopy, specifically ERCP. The potential risk of maternal pancreatitis affecting the fetus either iatrogenically or resulting from untreated choledocholithiasis is more significant. Of the 17 procedures we reported, 6 were performed safely during the first trimester: 2 for obstructive jaundice, 1 for gallstone pancreatitis, 1 for cholangitis, and 2 for suspected choledochal cysts. Given the hypothetical increased risk of spontaneous abortion during the first trimester, we stress that first trimester ERCP should only be performed for sound indications and when delaying intervention would offer more risk than benefit.

The exposure of radiation from fluoroscopy performed during ERCP is of particular concern given the potential risks to the fetus. According to the American College of Obstetricians and Gynecologists, exposure of less than 5 rad or 50 mGy does not appear to be associated with an appreciable increased rate of fetal anomalies or pregnancy loss.¹⁵ Fetal malformations, growth retardation, and intrauterine death have a threshold conceptus dose of 100–200 mGy.¹⁶ These doses are not typical for general diagnostic radiology, particularly when the fetus lies outside of the primary beam, as with ERCP.^17,18 Knowledge of radiation effects on fetal development is largely derived from epidemiologic and observational studies from exposed human populations (such as atomic bomb survivors) and animal studies.¹⁹

In prior studies, estimates of fetal radiation exposure have been quite variable, with published values ranging from 0.1 mGy to 3 mGy per procedure.^3,11,20 In a study by Kahaleh and associates, 15 patients underwent ERCP with thermoluminescent dosimetry (TLD) applied to the skin. Conceptus dosing was then estimated at 0.40 mGy (range, 0.01–1.8 mGy) using TLD readings.¹¹ TLD estimates of conceptus dosing are likely unreliable, as much of radiation exposure to the fetus may result from scatter radiation absorbed by the mother.¹⁷ Tham and colleagues used a nonanthropomorphic phantom to estimate the entrance skin dose and subsequently estimated fetal dose exposure at 3 mGy.³

Samara and colleagues recently presented an intriguing model utilizing data obtained from 24 nonpregnant patients for estimating conceptus radiation dosage for a specific patient procedure. The Monte-Carlo-N-particle code, a mathematical phantom used for simulating radiation dose distribution in the body, was then utilized to estimate the radiation dose delivered to a theoretical fetus. Their data revealed that fetal dose exposure may occasionally exceed 50 mGy (range, 3.4–55.9 mGy).¹⁷

These findings call special attention to dose-reduction techniques as well as methods for safe and effective fluoroscopy use, many of which have been outlined by Baron and Schueler in a recent editorial.¹⁸ These methods include limiting fluoroscopy time, using low-dose settings and the smallest possible field that is available, and using lead shielding between the patient and beam.¹⁸ In addition, patient positioning should be adjusted to allow for anterior-posterior beam projection when possible, as this positioning results in lower fetal dosing.^18,19 Alternative techniques to the use of fluoroscopy (such as wire-guided cannulation and bile aspiration methods as well as choledochoscopy to confirm stone clearance) should be utilized when feasible, though these techniques require additional expertise.¹⁸

Conclusion

Our data add to the limited number of existing reports demonstrating not only the safety of ERCP, but additionally the safety of therapeutic interventions such as stent placement, sphincterotomy, and balloon-stone extraction. We stress that these procedures be performed for sound indications with therapeutic intent and when nonintervention would potentially cause more harm. With careful attention to detail, these procedures can be performed safely by experienced endoscopists with judicious use of fluoroscopy.