Cardiovasc Intervent Radiol DOI 10.1007/s00270-015-1053-z
CLINICAL INVESTIGATION
Transjugular Intrahepatic Portosystemic Shunt Placement During Pregnancy: A Case Series of Five Patients Christopher R. Ingraham • Siddharth A. Padia • Guy E. Johnson • Thomas R. Easterling • Iris W. Liou • Kalpana M. Kanal • Karim Valji
Received: 20 September 2014 / Accepted: 25 December 2014 Ó Springer Science+Business Media New York and the Cardiovascular and Interventional Radiological Society of Europe (CIRSE) 2015
Abstract Background and Aims Complications of portal hypertension, such as variceal hemorrhage and ascites, are associated with significant increases in both mortality and complications during pregnancy. Transjugular intrahepatic portosystemic shunt (TIPS) is a well-established procedure for treating portal hypertension, but the safety of TIPS during pregnancy is largely unknown. In this series, we review five patients who underwent TIPS placement while pregnant and describe their clinical outcomes.
C. R. Ingraham (&) S. A. Padia G. E. Johnson K. Valji Department of Interventional Radiology, University of Washington, 1959 NE Pacific Street, Box 357115, Seattle, WA 98195-7115, USA e-mail: [email protected] S. A. Padia e-mail: [email protected] G. E. Johnson e-mail: [email protected] K. Valji e-mail: [email protected] T. R. Easterling Department of Obstetrics and Gynecology, University of Washington, Seattle, WA, USA e-mail: [email protected] I. W. Liou Department of Medicine, University of Washington, Seattle, WA, USA e-mail: [email protected] K. M. Kanal Physics Section, Department of Radiology, University of Washington, Seattle, WA, USA e-mail: [email protected]
Methods Five pregnant patients with cirrhosis and portal hypertension underwent elective TIPS for complications of portal hypertension (four for secondary prevention of variceal bleeding and one for refractory ascites). Outcomes measured were recurrent bleeding episodes or need for further paracenteses during pregnancy, estimated radiation dose to the fetus and gestational age at delivery. All patients were followed after delivery to evaluate technical and clinical success of the procedure. Results All five patients survived pregnancy and went on to deliver successfully. When TIPS was performed for secondary prevention of variceal bleeding (n = 4), no patients demonstrated variceal bleeding after TIPS placement. When TIPS was performed for refractory ascites (n = 1), no further paracenteses were required. All patients delivered successfully, albeit prematurely. Average radiation dose estimated to the fetus was 16.3 mGy. Conclusions This series suggests that TIPS can be performed in selective pregnant patients with portal hypertension, with little added risk to the mother or fetus. Keywords Portal hypertension Pregnancy Transjugular intrahepatic portosystemic shunt (TIPS)
Introduction In patients with cirrhosis, portal hypertension can manifest as ascites or bleeding esophageal varices. In women who are pregnant, cirrhosis and portal hypertension increase the risk of mortality during pregnancy, ranging from 10 to 61 % [1]. Bleeding of esophageal varices (EVs) during pregnancy is associated with mortality rates of up to 78 % [1].
123
C. R. Ingraham et al.: TIPS in Pregnancy
The mainstay for initial treatment of complications of portal hypertension includes treating the underlying cause of cirrhosis, salt restriction and the use of diuretics in the setting of ascites, and non-selective beta-blockers for variceal hemorrhage prophylaxis [2, 3]. If medical management fails, large-volume paracentesis can be performed for refractory ascites, and endoscopic variceal ligation can mitigate risk of variceal hemorrhage. In some patients who have failed conservative measures, transjugular intrahepatic portosystemic shunt (TIPS) placement may be an appropriate second-line treatment. By creating a TIPS, portal hypertension is either reduced or resolved, thereby reducing the risk of hemorrhage in patients with EVs, or reducing or eliminating ascites [4]. Although TIPS placement is well established in patients with complications of portal hypertension, it has not been thoroughly studied in pregnant patients. Potentially long procedure times raise concerns over radiation dose to the mother and fetus. However, the procedure has been described in just a few cases reports as salvage therapy for bleeding varices [5–7]. The purpose of this study was to describe our experience with TIPS placement in pregnant patients with severe portal hypertension: in four patients for prevention of variceal hemorrhage and in one patient with large volume, refractory ascites.
Materials and Methods A retrospective study was conducted at a single institution by querying the hospital’s patient information system to identify all cases of TIPS procedures performed in pregnant patients from January 2000 to July 2013. This study was approved by the hospital’s institutional review board (IRB). For this type of study, formal consent is not required. This group of patients was felt to be at high risk for portal hypertension-related complications during pregnancy. Prior to the TIPS procedure, all patients were seen in consultation by the interventional radiology (IR) service and counseled on the high-risk nature of the procedure and possible radiation risks to the fetus. The decision to perform the procedure was made jointly by the patient, IR, hepatology, and maternal-fetal-medicine teams. All patients were followed closely before and after delivery. TIPS procedures were performed by one of two interventional radiologists. All patients were under general anesthesia. One patient underwent TIPS placement using a Philips (Amsterdam, The Netherlands) angiography system and the remaining using a Siemens (Erlangen, Germany) system. After access into the right internal jugular vein was obtained, a long sheath was inserted and positioned in the hepatic vein. Needle passes within the hepatic
123
intraparenchymal tract were from the hepatic vein to attempt access into the portal vein. Frequently, several needle passes were necessary to successfully cannulate the portal vein. After gaining portal vein access, a portal venogram (Fig. 1) in a single projection was obtained to minimize radiation dose to the patient and fetus. Additionally, portal pressures in the portal vein and right atrium were obtained. After the sheath was advanced into the portal vein, a stent was deployed across the intraparenchymal tract and then balloon-dilated. In all cases, a Viatorr TIPS Endoprosthesis (Gore, Flagstaff, AZ, USA) was deployed. A final portal-systemic pressure gradient measurement and venogram (Fig. 2) were then obtained. In one patient, a large gastric varix was embolized post-TIPS placement using multiple fibered coils (Nester, Cook, Bloomington, IN, USA). To decrease fetal radiation dose, a lead apron was placed over the mother, shielding her lower abdomen/pelvis. Additionally, established methods [8] for dose reduction during the procedure were employed, notably using narrow collimation, intermittent or pulsed fluoroscopy, low-dose fluoroscopy settings, and avoidance of image magnification, when possible. A radiation dosimeter (thermoluminescent dosimeter, TLD) was placed at the level of the patient’s umbilicus, under the lead apron to record the radiation dose expected to reach the fetus during the procedure. The TLD was then sent to the manufacturer (Landauer, Glenwood, IL, USA) for measurement, and the fetal dose was estimated from the TLD reading. Use of a TLD placed at the pregnant mother’s umbilicus is a common method employed to estimate fetal dose during radiology procedures [9, 10]. The fetal dose is estimated to be
Fig. 1 TIPS placement in a pregnant patient. Portal venogram obtained after access was obtained into the anterior division of the right portal vein
4.3 No 7 13 No 8 17 weeks 3 days History of splenic artery aneurysm rupture Large esophageal varices despite prior splenectomy
22
37
4
5
Cirrhosis due to hepatitis B/D
10.1 No 1 14 No 7 24 weeks 6 days None Large esophageal and gastric varices
38 3
Cirrhosis due to congenital biliary atresia
5.6 No 5 22 Refractory 11 15 weeks 3 days None Refractory ascites
27 2
Alcoholic cirrhosis
54 Yes 3 10 No 9 18 weeks 2 days 4 History of recurrent variceal hemorrhage
7.7 No 4 12 No 8 16 weeks 30 1
Alcoholic cirrhosis
Indication for TIPS Etiology of liver disease Age (years) Patient
TIPS placement was performed in a total of five pregnant patients over approximately 4 years (Table 1). Four patients had known gastroesophageal varices. Two patients had prior variceal bleeding, while two were felt to be at high risk of bleeding due to large varices visualized during endoscopy. One of the five patients had refractory ascites, requiring large-volume paracenteses weekly prior to presentation. TIPS placement was successful in all five patients. Mean gestational age at which TIPS was performed was 18 weeks 3 days (range from 15 weeks 3 days to 24 weeks 6 days). The mean pre-TIPS porto-systemic pressure gradient was 14 mmHg (range of 10–22 mmHg), and the mean post-TIPS porto-systemic gradient was 4 mmHg (range of 1–7 mmHg). All patients survived pregnancy and went on to deliver live births.
Table 1 Demographic and procedural information
Results
4
Number of bleeding episodes prior to TIPS
about one-third of the entrance dose for the patient [11] and thus can be calculated from the TLD reading. Demographic and clinical parameters were recorded for each patient. Delivery details were recorded, and all patients were followed postpartum. Surveillance TIPS ultrasound was scheduled at 1 week, 1, 3, 6, and 12 months to assess patency of the stent. Frequency of ultrasound surveillance may have varied due to clinical status, patient adherence, and patient location.
History of recurrent variceal hemorrhage
Gestational age at MELD time of TIPS score prior (weeks) to TIPS
Fig. 2 TIPS placement in a pregnant patient. Venogram in the same patient as Fig. 1 after TIPS placement
Shistosomiasis
Ascites
Initial gradient (mmHg)
Final gradient (mmHg)
Varix embolization during TIPS
Estimated dose to fetus (mGy)
C. R. Ingraham et al.: TIPS in Pregnancy
123
C. R. Ingraham et al.: TIPS in Pregnancy Table 2 Delivery details Patient
Gestation age at delivery (weeks)
Reason for early delivery
1
34 weeks
Suspected placental abruption
2
32 weeks 1 day
Pre-eclampsia
3
27 weeks 5 days
4
31 weeks 1 day
Acute kidney injury, encephalopathy, DIC Pre-eclampsia
5
34 weeks 2 days
Non-reassuring fetal heart tracing and oligohydramnios
No episodes of variceal hemorrhage occurred in any of the patients after TIPS placement, either during the pregnancy or thereafter (follow-up ranged from 3 to 60 months, mean of 21 months). In the one patient with refractory ascites, no further paracenteses were required after TIPS creation. Mean gestational age at delivery was 31 weeks 6 days (range from 27 weeks 5 days to 34 weeks 2 days). Four patients delivered vaginally and one by cesarean section. Reasons for premature delivery are shown in Table 2. The TIPS remained patent throughout the remainder of the pregnancy in four of five patients. However, in one patient (patient 5), the TIPS required two revisions to treat two separate episodes of occlusion that occurred before delivery (occurring 1 week and 6 weeks after TIPS creation), attributed to the hypercoagulability of pregnancy and cirrhosis. This patient required therapeutic anti-coagulation for the duration of pregnancy and 8 weeks postpartum to maintain TIPS patency. Documented TIPS patency by ultrasound ranged from 3 to 60 months. One patient (patient 2) required a revision at 43 months after placement for a stenosis discovered on surveillance ultrasound. No patient experienced any of the major complications known to occur post-TIPS, including right heart failure, liver failure, or new or worsening hepatic encephalopathy. Patient 3 had baseline encephalopathy prior to the TIPS procedure that was well managed with lactulose. Post-TIPS the patient was continued on the same dose of lactulose without an exacerbation in her baseline encephalopathy. Around the time of her delivery (three months post-TIPS), she was non-compliant with lactulose, causing temporary encephalopathy. This patient also suffered from acute kidney injury prior to delivery, which was felt to be due to overdiuresis and not due to contrast administered during the TIPS procedure. One patient (patient 5) died of complications of progressive liver failure from cirrhosis and a non-variceal upper gastrointestinal hemorrhage 30 months after TIPS placement. The estimated absorbed radiation dose to the fetus ranged from 4.3 to 54.0 mGy (mean 16.3 mGy). It is
123
important to note that in the patient where the dose was estimated to be 54.0 mGy to the fetus (patient #2), access into the portal vein required several passes before successful access. Two overlapping Viatorr TIPS endoprostheses were required to adequately create the shunt. Moreover, this patient had four prior episodes of variceal bleeding before the TIPS procedure. After TIPS placement, persistent filling of a gastric varix was observed, which prompted the operator to embolize the varix during the procedure. All of these factors increased the fluoroscopy time and thus the dose to this patient and the fetus during the procedure.
Discussion Pregnancy in patients with advanced liver disease and portal hypertension is rare. Several case reports have described the TIPS procedure in pregnant patients [5–7]. However, in these reports, the emergent indication for TIPS placement was acute or recurrent variceal bleeding after failure of medical and/or endoscopic therapy. It has been proposed that the portal and esophageal venous systems have higher venous pressures during pregnancy in all patients [12]. During the second and third trimesters, there is increased circulating blood volume and compression of the IVC by the gravid uterus, particularly in the supine position. Theoretically, this causes increased shunting of blood to the lumbar, paraspinal, and azygous venous systems [12]. The esophageal veins drain into the azygous system, and increased pressure in this system would be expected to increase esophageal venous pressure. In non-cirrhotic pregnant patients, bleeding from transient esophageal varices may occur [13]. However, in pregnant patients with portal hypertension, the variceal vasculature is thought to be more susceptible to rupture [13]. Up to 78 % of varices will bleed during pregnancy, and the most frequently reported cause of maternal death in cirrhotic patients with portal hypertension is hemorrhage from EVs [14]. There are no formal guidelines on management of EVs in pregnancy. However, if no EVs are present prior to conception or if the status is unknown, some recommend endoscopy during the second trimester followed by ligation if significant varices are found [14–16]. Obviously, banding of EVs does not impact the risk of bleeding varices at other sites, such as gastric varices and pelvic or abdominal wall collaterals. The individuals with varices selected for TIPS in our series were all deemed to be at high risk for hemorrhage because of a history of recurrent bleeding EVs, EVs deemed too large to control with banding, or the presence of gastric varices.
C. R. Ingraham et al.: TIPS in Pregnancy
Controversy surrounds the use of nonselective betablockers for variceal hemorrhage prophylaxis. These medications are designated Food and Drug Administration pregnancy category C, due to concerns regarding fetal intrauterine growth retardation, neonatal bradycardia, and neonatal hypoglycemia [15, 16]. Moreover, the metabolism and clearance of beta-blockers are increased variably throughout the course of pregnancy, thus adequate dosing may be difficult to achieve or control [17, 18]. The use of medication prophylaxis alone against variceal bleeding was not thought to be a reliable method for these high-risk patients. Estimated absorbed radiation doses to the fetus were overall low in this patient population. This is expected given that the fluoroscopic field of view in the upper abdomen is away from the gravid uterus inferiorly. However, scatter radiation can reach the uterus, with a fetal dose estimated to be about one-third of the entrance exposure of the patient [11]. Therefore, diligent collimation and use of a lead apron over the lower abdomen further protect the fetus from radiation. If the fetal dose is of concern, a detailed calculation taking into account the fetal depth and patient size is conducted [19]. In this series, the highest estimated absorbed radiation dose to one fetus was 54 mGy (see results section for further description of this case). The other patients in our series had estimated fetal doses of approximately 10 mGy or less. Based on established guidelines from the American College of Radiology [20], no potential radiation effects are felt to occur or affect the fetus at doses less than 50 mGy at any gestational age. The fetus is felt to be most radiosensitive between 8 and 15 weeks, and potential effects are felt to occur at doses [100 mGy. The gestational age of the fetus in all pregnant patients in this series was greater than 15 weeks. In the pregnant patient who received an estimated absorbed dose of 54 mGy to the fetus, the procedure was performed at 18 weeks 2 days. Under the ACR guidelines, there are no potential radiation effects that are felt to occur in this gestational age range at this estimated dose. A recent review [10] states that there is no evidence demonstrating that radiation dose in or below diagnostic ranges (100 mGy) is associated with an increased incidence of congenital malformation, stillbirth, miscarriage, growth, or mental disability. Radiation effects to the conceptus are associated with doses greater than 100–200 mGy, with risks increasing at doses greater than 200 mGy. Therapeutic abortion is typically not recommended at doses less than 150 mGy [8, 20, 21]. Theoretical risks at and above this dose include cancer development, mental retardation and loss of IQ points,
and microcephaly [19, 22, 23]. It is important to note that these risks are dependent on the timing of the radiation exposure. Although these risks are higher early in the first trimester, theoretical risk extends into the early second trimester [20]. The development of cancer, the presence of mental retardation, or loss of IQ points can be difficult to quantify or identify in early childhood. Patients in this series have not been followed long enough to assess those markers. Additional maneuvers can be considered to lower potential fluoroscopy time (and therefore radiation dose) during the procedure in this patient population, such as transabdominal or intravascular ultrasound (IVUS) to assist with TIPS placement [24, 25]. While a single series did not show significant differences in fluoroscopy time in IVUSassisted TIPS compared to TIPS done via conventional methods [24], we believe that with increasing experience and familiarity using IVUS, the number of passes can be reduced. Additionally, the stent could also be potentially deployed under ultrasound guidance. By employing these additional strategies, the dose could be further reduced. While IVUS was not used for the TIPS procedures described in this series, we recently have adopted the routine use of IVUS to aid in portal vein puncture at our institution. This could be of further benefit to future pregnant patients requiring TIPS. This retrospective study has several limitations. The series was small, and pregnant patients with end-stage liver disease who did not undergo TIPS prior to delivery were not compared in this series. There is a potential for selection bias since the patients that were referred for TIPS by our hepatology service were felt to be suitable for the procedure and not likely to suffer from significant postprocedural complications. Although TIPS is a well-studied procedure with known benefits, it is not a benign procedure. Complications include bleeding (which in rare cases can be life-threatening), potential consequences of radiation exposure to the fetus, heart or liver failure, new or worsened hepatic encephalopathy, and TIPS dysfunction (thrombosis, occlusion, or failure). It is our hepatologists’ preference to manage the pregnant patient with portal hypertension medically and endoscopically, if possible, and reserve TIPS for only those patients with severe complications of portal hypertension. Another limitation is the estimation of fetal dose. Although use of a TLD placed at the umbilicus of the pregnant mother is a well-established method to estimate fetal dose, it is only an estimate. This method cannot measure radiation that may be reaching the fetus by scatter radiation. This method only estimates radiation that reaches the fetus via the direct beam.
123
C. R. Ingraham et al.: TIPS in Pregnancy
Conclusion Our data suggest that elective TIPS can be safely performed in the pregnant patient during the second trimester. Measured radiation doses in this series are below accepted thresholds. In selected patients, TIPS may reduce complications of portal hypertension, which can worsen during pregnancy, with little or no observed adverse effect on the fetus. Conflict of interest All authors declare that they have no financial conflicts of interest with this manuscript or its content. This study was not supported by grant funding. Statement of Informed Consent This study was approved by the hospital’s institutional review board (IRB). For this type of study, formal consent is not required. Statement of Human and Animal Rights All procedures performed in this study which involved human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
References 1. Russell MA, Craigo SD (1998) Cirrhosis and portal hypertension in pregnancy. Semin Perinatol 22:156–165 2. Leung W, Wong F (2011) Medical management of ascites. Expert Opin Pharmacother 12(8):1269–1283 3. Biecker E (2013) Portal hypertension and gastrointestinal bleeding: diagnosis, prevention and management. World J Gastroenterol 19(31):5035–5050 4. Fidelman N, Kwan SW, LaBerge JM et al (2012) The transjugular intrahepatic portosystemic shunt: an update. Am J Roentgenol 199(4):746–755 5. Wildberger JE, Vorwerk D, Winograd R et al (1998) New TIPS placement in pregnancy in recurrent esophageal varices hemorrhage—assessment of fetal radiation exposure. Rofo 169(4): 429–431 6. Savage C, Patel J, Lepe MR et al (2007) Transjugular intrahepatic portosystemic shunt creation for recurrent gastrointestinal bleeding during pregnancy. J Vasc Interv Radiol 18(7):902–904 7. Lodato F, Cappelli A, Montagnani M et al (2008) Transjugular intrahepatic portosystemic shunt: a case report of rescue management of unrestrainable variceal bleeding in a pregnant woman. Dig Liver Dis 40(5):387–390 8. Wieseler KM, Bhargava P, Kanal KM et al (2010) Imaging in pregnant patients: examination appropriateness. Radiographics 30(5):1215–1229 (discussion 1230–1233)
123
9. Hurwitz LM, Yoshizumi T, Reiman RE et al (2006) Radiation dose to the fetus from body MDCT during early gestation. Am J Roentgenol 186(3):871–876 10. Dauer LT, Thornton RH, Miller DL et al (2012) Radiation management for interventions using fluoroscopic or computed tomographic guidance during pregnancy: a joint guideline of the Society of Interventional Radiology and the Cardiovascular and Interventional Radiological Society of Europe with Endorsement by the Canadian Interventional Radiology Association. J Vasc Interv Radiol 23(1):19–32 11. Huda W (2009) Review of radiologic physics, 3rd edn. Lippincott Williams & Wilkins, Philadelphia 12. Homburg R, Bayer I, Lurie B (1988) Bleeding esophageal varices in pregnancy. A report of two cases. J Reprod Med 33:784–786 13. Britton RC (1982) Pregnancy and esophageal varices. Am J Surg 143:421–425 14. Zeeman GG, Moise JK (1999) Prophylactic banding of severe esophageal varices associated with liver cirrhosis in pregnancy. Obstet Gynecol 94:842 15. Tan J, Surti B, Saab S (2008) Pregnancy and cirrhosis. Liver Transpl 14:1081–1091 16. Hay JE (2008) Liver disease in pregnancy. Hepatology 47: 1067–1076 17. Hebert MF, Carr DB, Anderson GD et al (2005) Pharmacokinetics and pharmacodynamics of atenolol during pregnancy and postpartum. J Clin Pharmacol 45(1):25–33 18. Anderson GD, Carr DB (2009) Effect of pregnancy on the pharmacokinetics of antihypertensive drugs. Clin Pharmacokinet 48(3):159–168 19. Wagner LK, Lester RG, Saldana LR (1997) Exposure of the pregnant patient to diagnostic radiations: a guide to medical management, 2nd edn. Medical Physics, Madison 20. American College of Radiology (2008) ACR practice guideline for imaging pregnant or potentially pregnant adolescents and women with ionizing radiation. American College of Radiology, Reston 21. Martin JN Jr, Ridgway LE 3rd, Connors JJ et al (1990) Angiographic arterial embolization and computed tomography-directed drainage for the management of hemorrhage and infection with abdominal pregnancy. Obstet Gynecol 76(5 pt 2):941–945 22. Brent RL (2009) Saving lives and changing family histories: appropriate counseling of pregnant women and men and women of reproductive age, concerning the risk of diagnostic radiation exposures during and before pregnancy. Am J Obstet Gynecol 200(1):4–24 23. Bushberg JT (2002) The essential physics of medical imaging, 2nd edn, Lippincott Williams & Wilkins, Philadelphia 24. Farsad K, Fuss C, Kolbeck KJ et al (2012) Transjugular intrahepatic portosystemic shunt creation using intravascular ultrasound guidance. J Vasc Interv Radiol 23(12):1594–1602 25. Raza SA, Walser E, Hernandez A et al (2006) Transhepatic puncture of portal and hepatic veins for TIPS using a singleneedle pass under sonographic guidance. Am J Roentgenol 187(1):W87–W91
CLINICAL INVESTIGATION
Transjugular Intrahepatic Portosystemic Shunt Placement During Pregnancy: A Case Series of Five Patients Christopher R. Ingraham • Siddharth A. Padia • Guy E. Johnson • Thomas R. Easterling • Iris W. Liou • Kalpana M. Kanal • Karim Valji
Received: 20 September 2014 / Accepted: 25 December 2014 Ó Springer Science+Business Media New York and the Cardiovascular and Interventional Radiological Society of Europe (CIRSE) 2015
Abstract Background and Aims Complications of portal hypertension, such as variceal hemorrhage and ascites, are associated with significant increases in both mortality and complications during pregnancy. Transjugular intrahepatic portosystemic shunt (TIPS) is a well-established procedure for treating portal hypertension, but the safety of TIPS during pregnancy is largely unknown. In this series, we review five patients who underwent TIPS placement while pregnant and describe their clinical outcomes.
C. R. Ingraham (&) S. A. Padia G. E. Johnson K. Valji Department of Interventional Radiology, University of Washington, 1959 NE Pacific Street, Box 357115, Seattle, WA 98195-7115, USA e-mail: [email protected] S. A. Padia e-mail: [email protected] G. E. Johnson e-mail: [email protected] K. Valji e-mail: [email protected] T. R. Easterling Department of Obstetrics and Gynecology, University of Washington, Seattle, WA, USA e-mail: [email protected] I. W. Liou Department of Medicine, University of Washington, Seattle, WA, USA e-mail: [email protected] K. M. Kanal Physics Section, Department of Radiology, University of Washington, Seattle, WA, USA e-mail: [email protected]
Methods Five pregnant patients with cirrhosis and portal hypertension underwent elective TIPS for complications of portal hypertension (four for secondary prevention of variceal bleeding and one for refractory ascites). Outcomes measured were recurrent bleeding episodes or need for further paracenteses during pregnancy, estimated radiation dose to the fetus and gestational age at delivery. All patients were followed after delivery to evaluate technical and clinical success of the procedure. Results All five patients survived pregnancy and went on to deliver successfully. When TIPS was performed for secondary prevention of variceal bleeding (n = 4), no patients demonstrated variceal bleeding after TIPS placement. When TIPS was performed for refractory ascites (n = 1), no further paracenteses were required. All patients delivered successfully, albeit prematurely. Average radiation dose estimated to the fetus was 16.3 mGy. Conclusions This series suggests that TIPS can be performed in selective pregnant patients with portal hypertension, with little added risk to the mother or fetus. Keywords Portal hypertension Pregnancy Transjugular intrahepatic portosystemic shunt (TIPS)
Introduction In patients with cirrhosis, portal hypertension can manifest as ascites or bleeding esophageal varices. In women who are pregnant, cirrhosis and portal hypertension increase the risk of mortality during pregnancy, ranging from 10 to 61 % [1]. Bleeding of esophageal varices (EVs) during pregnancy is associated with mortality rates of up to 78 % [1].
123
C. R. Ingraham et al.: TIPS in Pregnancy
The mainstay for initial treatment of complications of portal hypertension includes treating the underlying cause of cirrhosis, salt restriction and the use of diuretics in the setting of ascites, and non-selective beta-blockers for variceal hemorrhage prophylaxis [2, 3]. If medical management fails, large-volume paracentesis can be performed for refractory ascites, and endoscopic variceal ligation can mitigate risk of variceal hemorrhage. In some patients who have failed conservative measures, transjugular intrahepatic portosystemic shunt (TIPS) placement may be an appropriate second-line treatment. By creating a TIPS, portal hypertension is either reduced or resolved, thereby reducing the risk of hemorrhage in patients with EVs, or reducing or eliminating ascites [4]. Although TIPS placement is well established in patients with complications of portal hypertension, it has not been thoroughly studied in pregnant patients. Potentially long procedure times raise concerns over radiation dose to the mother and fetus. However, the procedure has been described in just a few cases reports as salvage therapy for bleeding varices [5–7]. The purpose of this study was to describe our experience with TIPS placement in pregnant patients with severe portal hypertension: in four patients for prevention of variceal hemorrhage and in one patient with large volume, refractory ascites.
Materials and Methods A retrospective study was conducted at a single institution by querying the hospital’s patient information system to identify all cases of TIPS procedures performed in pregnant patients from January 2000 to July 2013. This study was approved by the hospital’s institutional review board (IRB). For this type of study, formal consent is not required. This group of patients was felt to be at high risk for portal hypertension-related complications during pregnancy. Prior to the TIPS procedure, all patients were seen in consultation by the interventional radiology (IR) service and counseled on the high-risk nature of the procedure and possible radiation risks to the fetus. The decision to perform the procedure was made jointly by the patient, IR, hepatology, and maternal-fetal-medicine teams. All patients were followed closely before and after delivery. TIPS procedures were performed by one of two interventional radiologists. All patients were under general anesthesia. One patient underwent TIPS placement using a Philips (Amsterdam, The Netherlands) angiography system and the remaining using a Siemens (Erlangen, Germany) system. After access into the right internal jugular vein was obtained, a long sheath was inserted and positioned in the hepatic vein. Needle passes within the hepatic
123
intraparenchymal tract were from the hepatic vein to attempt access into the portal vein. Frequently, several needle passes were necessary to successfully cannulate the portal vein. After gaining portal vein access, a portal venogram (Fig. 1) in a single projection was obtained to minimize radiation dose to the patient and fetus. Additionally, portal pressures in the portal vein and right atrium were obtained. After the sheath was advanced into the portal vein, a stent was deployed across the intraparenchymal tract and then balloon-dilated. In all cases, a Viatorr TIPS Endoprosthesis (Gore, Flagstaff, AZ, USA) was deployed. A final portal-systemic pressure gradient measurement and venogram (Fig. 2) were then obtained. In one patient, a large gastric varix was embolized post-TIPS placement using multiple fibered coils (Nester, Cook, Bloomington, IN, USA). To decrease fetal radiation dose, a lead apron was placed over the mother, shielding her lower abdomen/pelvis. Additionally, established methods [8] for dose reduction during the procedure were employed, notably using narrow collimation, intermittent or pulsed fluoroscopy, low-dose fluoroscopy settings, and avoidance of image magnification, when possible. A radiation dosimeter (thermoluminescent dosimeter, TLD) was placed at the level of the patient’s umbilicus, under the lead apron to record the radiation dose expected to reach the fetus during the procedure. The TLD was then sent to the manufacturer (Landauer, Glenwood, IL, USA) for measurement, and the fetal dose was estimated from the TLD reading. Use of a TLD placed at the pregnant mother’s umbilicus is a common method employed to estimate fetal dose during radiology procedures [9, 10]. The fetal dose is estimated to be
Fig. 1 TIPS placement in a pregnant patient. Portal venogram obtained after access was obtained into the anterior division of the right portal vein
4.3 No 7 13 No 8 17 weeks 3 days History of splenic artery aneurysm rupture Large esophageal varices despite prior splenectomy
22
37
4
5
Cirrhosis due to hepatitis B/D
10.1 No 1 14 No 7 24 weeks 6 days None Large esophageal and gastric varices
38 3
Cirrhosis due to congenital biliary atresia
5.6 No 5 22 Refractory 11 15 weeks 3 days None Refractory ascites
27 2
Alcoholic cirrhosis
54 Yes 3 10 No 9 18 weeks 2 days 4 History of recurrent variceal hemorrhage
7.7 No 4 12 No 8 16 weeks 30 1
Alcoholic cirrhosis
Indication for TIPS Etiology of liver disease Age (years) Patient
TIPS placement was performed in a total of five pregnant patients over approximately 4 years (Table 1). Four patients had known gastroesophageal varices. Two patients had prior variceal bleeding, while two were felt to be at high risk of bleeding due to large varices visualized during endoscopy. One of the five patients had refractory ascites, requiring large-volume paracenteses weekly prior to presentation. TIPS placement was successful in all five patients. Mean gestational age at which TIPS was performed was 18 weeks 3 days (range from 15 weeks 3 days to 24 weeks 6 days). The mean pre-TIPS porto-systemic pressure gradient was 14 mmHg (range of 10–22 mmHg), and the mean post-TIPS porto-systemic gradient was 4 mmHg (range of 1–7 mmHg). All patients survived pregnancy and went on to deliver live births.
Table 1 Demographic and procedural information
Results
4
Number of bleeding episodes prior to TIPS
about one-third of the entrance dose for the patient [11] and thus can be calculated from the TLD reading. Demographic and clinical parameters were recorded for each patient. Delivery details were recorded, and all patients were followed postpartum. Surveillance TIPS ultrasound was scheduled at 1 week, 1, 3, 6, and 12 months to assess patency of the stent. Frequency of ultrasound surveillance may have varied due to clinical status, patient adherence, and patient location.
History of recurrent variceal hemorrhage
Gestational age at MELD time of TIPS score prior (weeks) to TIPS
Fig. 2 TIPS placement in a pregnant patient. Venogram in the same patient as Fig. 1 after TIPS placement
Shistosomiasis
Ascites
Initial gradient (mmHg)
Final gradient (mmHg)
Varix embolization during TIPS
Estimated dose to fetus (mGy)
C. R. Ingraham et al.: TIPS in Pregnancy
123
C. R. Ingraham et al.: TIPS in Pregnancy Table 2 Delivery details Patient
Gestation age at delivery (weeks)
Reason for early delivery
1
34 weeks
Suspected placental abruption
2
32 weeks 1 day
Pre-eclampsia
3
27 weeks 5 days
4
31 weeks 1 day
Acute kidney injury, encephalopathy, DIC Pre-eclampsia
5
34 weeks 2 days
Non-reassuring fetal heart tracing and oligohydramnios
No episodes of variceal hemorrhage occurred in any of the patients after TIPS placement, either during the pregnancy or thereafter (follow-up ranged from 3 to 60 months, mean of 21 months). In the one patient with refractory ascites, no further paracenteses were required after TIPS creation. Mean gestational age at delivery was 31 weeks 6 days (range from 27 weeks 5 days to 34 weeks 2 days). Four patients delivered vaginally and one by cesarean section. Reasons for premature delivery are shown in Table 2. The TIPS remained patent throughout the remainder of the pregnancy in four of five patients. However, in one patient (patient 5), the TIPS required two revisions to treat two separate episodes of occlusion that occurred before delivery (occurring 1 week and 6 weeks after TIPS creation), attributed to the hypercoagulability of pregnancy and cirrhosis. This patient required therapeutic anti-coagulation for the duration of pregnancy and 8 weeks postpartum to maintain TIPS patency. Documented TIPS patency by ultrasound ranged from 3 to 60 months. One patient (patient 2) required a revision at 43 months after placement for a stenosis discovered on surveillance ultrasound. No patient experienced any of the major complications known to occur post-TIPS, including right heart failure, liver failure, or new or worsening hepatic encephalopathy. Patient 3 had baseline encephalopathy prior to the TIPS procedure that was well managed with lactulose. Post-TIPS the patient was continued on the same dose of lactulose without an exacerbation in her baseline encephalopathy. Around the time of her delivery (three months post-TIPS), she was non-compliant with lactulose, causing temporary encephalopathy. This patient also suffered from acute kidney injury prior to delivery, which was felt to be due to overdiuresis and not due to contrast administered during the TIPS procedure. One patient (patient 5) died of complications of progressive liver failure from cirrhosis and a non-variceal upper gastrointestinal hemorrhage 30 months after TIPS placement. The estimated absorbed radiation dose to the fetus ranged from 4.3 to 54.0 mGy (mean 16.3 mGy). It is
123
important to note that in the patient where the dose was estimated to be 54.0 mGy to the fetus (patient #2), access into the portal vein required several passes before successful access. Two overlapping Viatorr TIPS endoprostheses were required to adequately create the shunt. Moreover, this patient had four prior episodes of variceal bleeding before the TIPS procedure. After TIPS placement, persistent filling of a gastric varix was observed, which prompted the operator to embolize the varix during the procedure. All of these factors increased the fluoroscopy time and thus the dose to this patient and the fetus during the procedure.
Discussion Pregnancy in patients with advanced liver disease and portal hypertension is rare. Several case reports have described the TIPS procedure in pregnant patients [5–7]. However, in these reports, the emergent indication for TIPS placement was acute or recurrent variceal bleeding after failure of medical and/or endoscopic therapy. It has been proposed that the portal and esophageal venous systems have higher venous pressures during pregnancy in all patients [12]. During the second and third trimesters, there is increased circulating blood volume and compression of the IVC by the gravid uterus, particularly in the supine position. Theoretically, this causes increased shunting of blood to the lumbar, paraspinal, and azygous venous systems [12]. The esophageal veins drain into the azygous system, and increased pressure in this system would be expected to increase esophageal venous pressure. In non-cirrhotic pregnant patients, bleeding from transient esophageal varices may occur [13]. However, in pregnant patients with portal hypertension, the variceal vasculature is thought to be more susceptible to rupture [13]. Up to 78 % of varices will bleed during pregnancy, and the most frequently reported cause of maternal death in cirrhotic patients with portal hypertension is hemorrhage from EVs [14]. There are no formal guidelines on management of EVs in pregnancy. However, if no EVs are present prior to conception or if the status is unknown, some recommend endoscopy during the second trimester followed by ligation if significant varices are found [14–16]. Obviously, banding of EVs does not impact the risk of bleeding varices at other sites, such as gastric varices and pelvic or abdominal wall collaterals. The individuals with varices selected for TIPS in our series were all deemed to be at high risk for hemorrhage because of a history of recurrent bleeding EVs, EVs deemed too large to control with banding, or the presence of gastric varices.
C. R. Ingraham et al.: TIPS in Pregnancy
Controversy surrounds the use of nonselective betablockers for variceal hemorrhage prophylaxis. These medications are designated Food and Drug Administration pregnancy category C, due to concerns regarding fetal intrauterine growth retardation, neonatal bradycardia, and neonatal hypoglycemia [15, 16]. Moreover, the metabolism and clearance of beta-blockers are increased variably throughout the course of pregnancy, thus adequate dosing may be difficult to achieve or control [17, 18]. The use of medication prophylaxis alone against variceal bleeding was not thought to be a reliable method for these high-risk patients. Estimated absorbed radiation doses to the fetus were overall low in this patient population. This is expected given that the fluoroscopic field of view in the upper abdomen is away from the gravid uterus inferiorly. However, scatter radiation can reach the uterus, with a fetal dose estimated to be about one-third of the entrance exposure of the patient [11]. Therefore, diligent collimation and use of a lead apron over the lower abdomen further protect the fetus from radiation. If the fetal dose is of concern, a detailed calculation taking into account the fetal depth and patient size is conducted [19]. In this series, the highest estimated absorbed radiation dose to one fetus was 54 mGy (see results section for further description of this case). The other patients in our series had estimated fetal doses of approximately 10 mGy or less. Based on established guidelines from the American College of Radiology [20], no potential radiation effects are felt to occur or affect the fetus at doses less than 50 mGy at any gestational age. The fetus is felt to be most radiosensitive between 8 and 15 weeks, and potential effects are felt to occur at doses [100 mGy. The gestational age of the fetus in all pregnant patients in this series was greater than 15 weeks. In the pregnant patient who received an estimated absorbed dose of 54 mGy to the fetus, the procedure was performed at 18 weeks 2 days. Under the ACR guidelines, there are no potential radiation effects that are felt to occur in this gestational age range at this estimated dose. A recent review [10] states that there is no evidence demonstrating that radiation dose in or below diagnostic ranges (100 mGy) is associated with an increased incidence of congenital malformation, stillbirth, miscarriage, growth, or mental disability. Radiation effects to the conceptus are associated with doses greater than 100–200 mGy, with risks increasing at doses greater than 200 mGy. Therapeutic abortion is typically not recommended at doses less than 150 mGy [8, 20, 21]. Theoretical risks at and above this dose include cancer development, mental retardation and loss of IQ points,
and microcephaly [19, 22, 23]. It is important to note that these risks are dependent on the timing of the radiation exposure. Although these risks are higher early in the first trimester, theoretical risk extends into the early second trimester [20]. The development of cancer, the presence of mental retardation, or loss of IQ points can be difficult to quantify or identify in early childhood. Patients in this series have not been followed long enough to assess those markers. Additional maneuvers can be considered to lower potential fluoroscopy time (and therefore radiation dose) during the procedure in this patient population, such as transabdominal or intravascular ultrasound (IVUS) to assist with TIPS placement [24, 25]. While a single series did not show significant differences in fluoroscopy time in IVUSassisted TIPS compared to TIPS done via conventional methods [24], we believe that with increasing experience and familiarity using IVUS, the number of passes can be reduced. Additionally, the stent could also be potentially deployed under ultrasound guidance. By employing these additional strategies, the dose could be further reduced. While IVUS was not used for the TIPS procedures described in this series, we recently have adopted the routine use of IVUS to aid in portal vein puncture at our institution. This could be of further benefit to future pregnant patients requiring TIPS. This retrospective study has several limitations. The series was small, and pregnant patients with end-stage liver disease who did not undergo TIPS prior to delivery were not compared in this series. There is a potential for selection bias since the patients that were referred for TIPS by our hepatology service were felt to be suitable for the procedure and not likely to suffer from significant postprocedural complications. Although TIPS is a well-studied procedure with known benefits, it is not a benign procedure. Complications include bleeding (which in rare cases can be life-threatening), potential consequences of radiation exposure to the fetus, heart or liver failure, new or worsened hepatic encephalopathy, and TIPS dysfunction (thrombosis, occlusion, or failure). It is our hepatologists’ preference to manage the pregnant patient with portal hypertension medically and endoscopically, if possible, and reserve TIPS for only those patients with severe complications of portal hypertension. Another limitation is the estimation of fetal dose. Although use of a TLD placed at the umbilicus of the pregnant mother is a well-established method to estimate fetal dose, it is only an estimate. This method cannot measure radiation that may be reaching the fetus by scatter radiation. This method only estimates radiation that reaches the fetus via the direct beam.
123
C. R. Ingraham et al.: TIPS in Pregnancy
Conclusion Our data suggest that elective TIPS can be safely performed in the pregnant patient during the second trimester. Measured radiation doses in this series are below accepted thresholds. In selected patients, TIPS may reduce complications of portal hypertension, which can worsen during pregnancy, with little or no observed adverse effect on the fetus. Conflict of interest All authors declare that they have no financial conflicts of interest with this manuscript or its content. This study was not supported by grant funding. Statement of Informed Consent This study was approved by the hospital’s institutional review board (IRB). For this type of study, formal consent is not required. Statement of Human and Animal Rights All procedures performed in this study which involved human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
References 1. Russell MA, Craigo SD (1998) Cirrhosis and portal hypertension in pregnancy. Semin Perinatol 22:156–165 2. Leung W, Wong F (2011) Medical management of ascites. Expert Opin Pharmacother 12(8):1269–1283 3. Biecker E (2013) Portal hypertension and gastrointestinal bleeding: diagnosis, prevention and management. World J Gastroenterol 19(31):5035–5050 4. Fidelman N, Kwan SW, LaBerge JM et al (2012) The transjugular intrahepatic portosystemic shunt: an update. Am J Roentgenol 199(4):746–755 5. Wildberger JE, Vorwerk D, Winograd R et al (1998) New TIPS placement in pregnancy in recurrent esophageal varices hemorrhage—assessment of fetal radiation exposure. Rofo 169(4): 429–431 6. Savage C, Patel J, Lepe MR et al (2007) Transjugular intrahepatic portosystemic shunt creation for recurrent gastrointestinal bleeding during pregnancy. J Vasc Interv Radiol 18(7):902–904 7. Lodato F, Cappelli A, Montagnani M et al (2008) Transjugular intrahepatic portosystemic shunt: a case report of rescue management of unrestrainable variceal bleeding in a pregnant woman. Dig Liver Dis 40(5):387–390 8. Wieseler KM, Bhargava P, Kanal KM et al (2010) Imaging in pregnant patients: examination appropriateness. Radiographics 30(5):1215–1229 (discussion 1230–1233)
123
9. Hurwitz LM, Yoshizumi T, Reiman RE et al (2006) Radiation dose to the fetus from body MDCT during early gestation. Am J Roentgenol 186(3):871–876 10. Dauer LT, Thornton RH, Miller DL et al (2012) Radiation management for interventions using fluoroscopic or computed tomographic guidance during pregnancy: a joint guideline of the Society of Interventional Radiology and the Cardiovascular and Interventional Radiological Society of Europe with Endorsement by the Canadian Interventional Radiology Association. J Vasc Interv Radiol 23(1):19–32 11. Huda W (2009) Review of radiologic physics, 3rd edn. Lippincott Williams & Wilkins, Philadelphia 12. Homburg R, Bayer I, Lurie B (1988) Bleeding esophageal varices in pregnancy. A report of two cases. J Reprod Med 33:784–786 13. Britton RC (1982) Pregnancy and esophageal varices. Am J Surg 143:421–425 14. Zeeman GG, Moise JK (1999) Prophylactic banding of severe esophageal varices associated with liver cirrhosis in pregnancy. Obstet Gynecol 94:842 15. Tan J, Surti B, Saab S (2008) Pregnancy and cirrhosis. Liver Transpl 14:1081–1091 16. Hay JE (2008) Liver disease in pregnancy. Hepatology 47: 1067–1076 17. Hebert MF, Carr DB, Anderson GD et al (2005) Pharmacokinetics and pharmacodynamics of atenolol during pregnancy and postpartum. J Clin Pharmacol 45(1):25–33 18. Anderson GD, Carr DB (2009) Effect of pregnancy on the pharmacokinetics of antihypertensive drugs. Clin Pharmacokinet 48(3):159–168 19. Wagner LK, Lester RG, Saldana LR (1997) Exposure of the pregnant patient to diagnostic radiations: a guide to medical management, 2nd edn. Medical Physics, Madison 20. American College of Radiology (2008) ACR practice guideline for imaging pregnant or potentially pregnant adolescents and women with ionizing radiation. American College of Radiology, Reston 21. Martin JN Jr, Ridgway LE 3rd, Connors JJ et al (1990) Angiographic arterial embolization and computed tomography-directed drainage for the management of hemorrhage and infection with abdominal pregnancy. Obstet Gynecol 76(5 pt 2):941–945 22. Brent RL (2009) Saving lives and changing family histories: appropriate counseling of pregnant women and men and women of reproductive age, concerning the risk of diagnostic radiation exposures during and before pregnancy. Am J Obstet Gynecol 200(1):4–24 23. Bushberg JT (2002) The essential physics of medical imaging, 2nd edn, Lippincott Williams & Wilkins, Philadelphia 24. Farsad K, Fuss C, Kolbeck KJ et al (2012) Transjugular intrahepatic portosystemic shunt creation using intravascular ultrasound guidance. J Vasc Interv Radiol 23(12):1594–1602 25. Raza SA, Walser E, Hernandez A et al (2006) Transhepatic puncture of portal and hepatic veins for TIPS using a singleneedle pass under sonographic guidance. Am J Roentgenol 187(1):W87–W91
