Four-dimensional flow magnetic resonance imaging for assessment of hemodynamic changes in the portal venous system before and after balloon-occluded retrograde transvenous obliteration: a pilot feasibility study

Abstract

Background

The effectiveness of four-dimensional (4D) flow magnetic resonance imaging (MRI) for assessing hemodynamic changes before and after balloon-occluded retrograde transvenous obliteration (BRTO) remains unclear.

Purpose

To evaluate the feasibility of 4D flow MRI for assessing hemodynamic changes in the portal venous system before and after BRTO.

Material and Methods

We included 10 patients (7 men, 3 women; mean age = 67 years) with liver cirrhosis who had a high risk of gastric variceal bleeding or hepatic encephalopathy. Non-contrast 4D flow MRI of the upper abdomen was performed before and after BRTO. In addition, we compared the blood flow rates in the portal vein (PV), superior mesenteric vein (SMV), splenic vein (SV), left renal vein, and inferior vena cava before and after BRTO. Moreover, the flow directions of the SMV and SV before and after BRTO were assessed using both portography and 4D flow MRI.

Results

There was a significant post-BRTO increase in the blood flow rate in the PV and SV (P < 0.05). There was no significant post-BRTO change in the blood flow rates in the SMV, inferior vena cava, and left renal vein. In four patients, portography confirmed that hepatofugal flow in the SV and SMV changed to hepatopetal flow after BRTO. Moreover, 4D flow MRI correctly assessed the flow directions in the SMV and SV in 70%–100% of the patients.

Conclusion

4D flow MRI can be used to detect hemodynamic changes in the portal venous system before and after BRTO.

Keywords

Four-dimensional flow magnetic resonance imaging portal vein balloon-occluded retrograde transvenous obliteration

Introduction

Gastric and esophageal variceal bleeding and hepatic encephalopathy are serious and life-threatening complications in patients with liver cirrhosis or non-cirrhotic portal hypertension (1). In these patients, interventions such as a transjugular intrahepatic portosystemic shunt can be used to relieve portal hypertension (2,3). Balloon-occluded retrograde transvenous obliteration (BRTO) (4 –6) is the first-line treatment for gastric varices involving an increased hemorrhagic risk (7 –9); in addition, assessing portal hemodynamics can help determine indications for BRTO. Ultrasonography is a simple and useful procedure for hemodynamic assessment; however, it involves considerable operator dependence and limited accuracy in the measurement of blood flow rate (10 –12). Contrast-enhanced computed tomography (CT) allows precise morphological evaluation of gastroesophageal varices and splenorenal shunts with wide image coverage. However, it provides limited information regarding the blood flow rate and direction in the portal venous system. Moreover, angiography is the gold standard for hemodynamic evaluation of the portal venous system (13) and provides information regarding blood flow direction. However, it is an invasive procedure involving radiation exposure; furthermore, it cannot quantify the blood flow rate and velocity.

Four-dimensional (4D) flow magnetic resonance imaging (MRI) is a feasible non-invasive test for hemodynamic assessment (14) and is mainly used to investigate patients with cardiovascular and cerebrovascular diseases (15,16). Recent studies have demonstrated the clinical utility of 4D flow MRI in the abdominal region (17 –19), including assessment of the portal venous system (20 –22). Hyodo et al. demonstrated the feasibility of 4D flow MRI for evaluating portal venous blood flow before and after shunt embolization in patients with portosystemic shunt-related hepatic encephalopathy (20). We speculated that 4D flow MRI, which facilitates quantitative blood flow measurement and qualitative flow direction, can detect dynamic changes in portal venous system circulation before and after BRTO. However, the effectiveness of 4D flow MRI in assessing hemodynamic changes before and after BRTO remains unclear.

Therefore, the aim of the present study was to evaluate the utility of 4D flow MRI in assessing hemodynamic changes in the portal venous system before and after BRTO.

Material and Methods

Participants

This prospective study evaluated the feasibility of 4D flow MRI for detecting hemodynamic changes in the portal venous system before and after BRTO in patients with gastric varices. This study was approved by the ethics review board of our institution (approval protocol no. 2287); further, all patients provided written informed consent.

We consecutively enrolled 10 patients (7 men, 3 women) with hepatic cirrhosis who had gastric varices and splenic or gastro-renal shunt and were scheduled to undergo BRTO between March 2020 and May 2021. 4D flow MRI of the portal venous system was scheduled within one week before and after BRTO. After the inclusion of the 10 patients, patient enrolment was stopped given the infrequent patient referral and predefined study period. During BRTO, a balloon catheter was inserted into the gastro-renal or gastro-caval shunt to block the shunt flow using an inflated balloon. Under temporary balloon occlusion, an embolic agent (monoethanolamine oleate with/without platinum-based coils) was injected through the catheter into the abnormally dilated vein to obliterate the shunt. The pressure gradient of the main portal vein was measured before and after obliteration using a wedge catheter in the hepatic vein. Table 1 summarizes the patients’ characteristics, Child-Pugh grade before and after BRTO, and pressure gradient of the main portal vein during BRTO.

Table 1.
Patient characteristics, clinical liver function, and measured pressure gradient before and after BRTO.

Parameters Values

No. of participants 10

Sex

Male 7

Female 3

Age (years) 67 (58–79)

Body weight (kg) 71 (45–101)

Etiology of chronic liver disease

Hepatitis B 1

Hepatitis C 4

Alcohol consumption 3

NASH 2

Indication of BRTO

High risk of gastric variceal bleeding 7

Hepatic encephalopathy 3

Child-Pugh score before BRTO

Class A 5

Class B 3

Class C 2

Child-Pugh score after 3 months of BRTO

Class A 8

Class B 2

Pressure gradient (mmHg)

Before BRTO 15.4 (5–22)

After BRTO 20.4 (11–33)

Parameters	Values
No. of participants	10
Sex
Male	7
Female	3
Age (years)	67 (58–79)
Body weight (kg)	71 (45–101)
Etiology of chronic liver disease
Hepatitis B	1
Hepatitis C	4
Alcohol consumption	3
NASH	2
Indication of BRTO
High risk of gastric variceal bleeding	7
Hepatic encephalopathy	3
Child-Pugh score before BRTO
Class A	5
Class B	3
Class C	2
Child-Pugh score after 3 months of BRTO
Class A	8
Class B	2
Pressure gradient (mmHg)
Before BRTO	15.4 (5–22)
After BRTO	20.4 (11–33)

Values are given as n or mean (range).

BRTO, balloon-occluded retrograde transvenous obliteration; NASH, non-alcoholic steatohepatitis.

4D flow MRI

We used a clinical 3-T scanner (SIGNA Premier; GE Healthcare, Chicago, IL, USA) with a 30-channel phased-array body coil (AIR AA Coil; GE Healthcare, Chicago, IL, USA). We performed a 4D flow MRI without contrast agents. All patients fasted for 4 h before imaging to prevent hemodynamic changes in the portal venous system (23). 4D velocity mapping was conducted to cover the upper abdomen using a three-dimensional phase-contrast technique under free-breathing and electrocardiogram synchronization. The MRI parameters were as follows: imaging volume = 40 × 40 × 25 cm; spatial resolution = 1.6 × 1.6 × 2.2 mm; number of slabs = 3; overlap = 10 mm; TR/TE = 4.9/2.7 ms; flip angle = 8°; and velocity-encoding sensitivity = 60 cm/s. All scans were reconstructed using a sparse sampling model (Hyperkat; GE Healthcare, Chicago, IL, USA) to reduce the imaging time (24). Hyperkat combines parallel imaging and view sharing; further, it applies spatial and temporal correlations of MRI data to improve the temporal resolution. The total imaging time depended on the heart rate and was approximately 20 min for each patient. Table 2 shows the imaging parameters of 4D flow MRI.

Table 2.

Four-dimensional flow MRI parameters.

Parameters	Values
Field of view (cm³)	40 × 40 × 25
Spatial resolution (mm³)	1.6 × 1.6 × 2.2
Repetition time (ms)	4.9
Echo time (ms)	2.7
Velocity encoding sensitivity (cm/s)	60
Flip angle (°)	8
Band width (kHz/field of view)	62.5
Temporal resolution (ms)	63
No. of phases	20
No. of slabs	3
Fat suppression	None

Quantitative and qualitative image analysis

The acquired 4D flow MRI data were reconstructed for 20 timeframes per cardiac cycle. The phase offsets for Maxwell terms and eddy currents were automatically corrected during reconstruction. Moreover, velocity-weighted angiograms were calculated from the final velocity and magnitude data for all 20 timeframes. A vendor-supplied algorithm was used for background phase correction. Data postprocessing was performed using the iTFlow software (Cardio Flow Design, Tokyo, Japan), which provides 4D flow dynamics of the portal venous system (Fig. 1a). We measured blood flow and velocity in the following five locations of the portal venous system (Fig. 1b): (i) portal vein (PV); (ii) splenic vein (SV); (iii) superior mesenteric vein (SMV); (iv) left renal vein (LRV); and (v) inferior vena cava (IVC). The blood flow rate (L/min) was measured in each segment by two observers (MRI technologists with 7 and 10 years of experience in phase-contrast MRI who were blinded to the clinical information, BRTO timing, and portography results) independently by manually placing slices cut perpendicular to the vessel long axis on the clearly delineated vessel segment. Both observers could access pretreatment contrast-enhanced CT images of all patients to facilitate detection of the portal venous system. We excluded non-measurable segments of 4D flow MRI scans due to slow blood flow or small vessel calibers, including three cases at the PV and SV, two cases at the SMV, and one case at the LRV, from the subsequent quantitative analysis. One observer assessed the amount of ascites on the magnitude 4D-flow MR image or routine abdominal MR images (e.g. T2-weighted image) using the following rating system: no, small, moderate, and large amounts of ascites.

Two interventional radiologists with 17 and 20 years of experience who were blinded to the pretreatment 4D flow MRI findings assessed the flow direction of the SV and SMV by consensus on both 4D flow MRI and portography. Flow direction was classified as hepatopetal (antegrade flow toward the liver), hepatofugal (retrograde flow toward the spleen or small bowel), or undetectable.

Statistical analysis

Intraclass correlation coefficient (ICC) analysis was performed to assess agreement between both observers’ measurements. Differences in the blood flow of the portal venous system and pressure gradient before and after BRTO were analyzed using the Wilcoxon signed-rank test. The percentage of detectable blood flow direction on 4D flow MRI was described using the flow direction confirmed by portography during BRTO as a reference. Statistical significance was set at P < 0.05. Statistical analyses were performed using MATLAB (MathWorks Inc., Natick, MA, USA) and Prism 9 (GraphPad Software, San Diego, CA, USA).

Results

All 10 included patients underwent 4D flow MRI before and after BRTO. Table 1 summarizes the patients’ demographics. The pressure gradient of the main portal vein was significantly higher after BRTO than before BRTO (P < 0.05) (Table 1).

The intra-observer ICC of the blood flow measurement was 0.90 (95% confidence interval = 0.81–0.98), which indicated excellent agreement. There was a significant post-BRTO increase in the blood flow rate in the PV and SV (P < 0.05) (Table 3), but not in the SMV, LRV, and IVC (P > 0.05). Table 3 and Fig. 2 show the results of the blood flow rate measured on 4D flow MRI before and after BRTO of each segment. Regarding the per-patient-based blood flow rate changes, all patients showed a post-BRTO increase in the blood flow rate in the PV and SV as well as a post-BRTO decrease in the LRV. Contrastingly, one patient showed an increased blood flow rate after BRTO in the SMV and IVC. A representative case is shown in Fig. 3, which demonstrated a change from hepatofugal blood flow in the SV and SMV before BRTO to hepatopetal blood flow after BRTO. In four patients, portography confirmed a change from hepatofugal blood flow in four patients in the SV and SMV before BRTO to hepatopetal blood flow after BRTO. 4D flow MRI could correctly detect the blood flow direction in 100% (10/10) and 90% (9/10) of the patients in the SV and SMV, respectively, before BRTO, as well as 70% (7/10) and 80% (8/10) of the patients in the SV and SMV, respectively, after BRTO. Undetectable blood flow direction could be attributed to slow blood flow or small vessel calibers. There were no disagreements in the blood flow direction between portography and 4D flow MRI. Before BRTO, there were nine patients without detectable ascites and one patient with a small amount of ascites. Contrastingly, after BRTO, there were seven patients without detectable ascites and three patients with a small amount of ascites after BRTO. The amount of ascites increased and remained unchanged in two and eight patients, respectively.

Table 3.

Blood flow rate in the portal venous system measured by 4D flow MRI before and after BRTO.

	Flow rate before BRTO (L/min)	Flow rate after BRTO (L/min)	P value*
Observer 1
PV	0.44 ± 0.32	0.81 ± 0.29	0.02
SV	0.02 ± 0.31	0.38 ± 0.26	<0.01
SMV	0.42 ± 0.16	0.29 ± 0.14	0.09
Left RV	0.79 ± 0.47	0.50 ± 0.27	0.09
IVC	3.52 ± 1.22	2.81 ± 0.77	0.14
Observer 2
PV	0.44 ± 0.33	0.73 ± 0.28	0.01
SV	0.02 ± 0.32	0.41 ± 0.28	<0.01
SMV	0.43 ± 0.17	0.29 ± 0.13	0.11
Left RV	0.77 ± 0.49	0.51 ± 0.31	0.15
IVC	3.54 ± 1.20	2.79 ± 0.74	0.10

Values are given as mean ± standard deviation.

*Differences in the blood flow rate in the portal venous system before and after BRTO (paired t-test).

BRTO, balloon-occluded retrograde transvenous obliteration; IVC, inferior vena cava; LRV, left renal vein; PV, portal vein; SMV, superior mesenteric vein; SV, splenic vein.

Discussion

In this prospective study, 4D flow MRI successfully revealed hemodynamic changes in the portal venous system before and after BRTO in patients with gastric varices. Specifically, there was a significantly increased blood flow rate in the PV and SV as well as a dramatic change in the flow direction in the portal venous system after BRTO.

Quantitative assessment using 4D flow MRI revealed increased blood flow in the PV and SV as well as reduced blood flow in the IVC, SMV, and LRV after BRTO. Blood outflow from the portal venous system to the systemic circulation through the splenorenal shunt and gastric varices can be blocked through BRTO, which restores the natural anterograde blood flow in the SV and PV toward the liver and decreases blood flow into the LRV and IVC. Notably, there was a post-BRTO decrease in the blood flow rate in the SMV. Increased blood flow in the SV after BRTO could have led to increased blood flow and pressure in the PV (25,26). Increased blood pressure in the PV could affect blood flow velocity and reduce the blood flow rate in the SMV.

The blood flow rate measured on 4D flow MRI is primarily affected by the blood vessel diameter and blood flow velocity, which should be considered when interpreting the findings. When the blood flow velocity is outside the range of the velocity-encoding sensitivity (25%–75% of the velocity-encoding sensitivity), it could be underestimated, which leads to a decreased blood flow rate (27). In our study, this might have happened given the low blood flow velocity, especially in the PV and SV before BRTO. Accordingly, it may be useful to apply velocity-encoding sensitivity adjustment and other techniques, including multiple velocity-encoding sensitivity (28,29).

In our study, 4D flow MRI showed a high, but not perfect, detection rate of blood flow direction. As aforementioned, slow blood flow outside the range of velocity-encoding sensitivity may result in an undetectable flow signal in the pathline 4D flow MR image. In our study, this could be demonstrated by the post-BRTO detection rate in the SV (70%). This could be attributed to the change from stable hepatofugal blood flow in the SV before BRTO to unstable hepatopetal blood flow after BRTO. In the other cases, there was consistency between the blood flow direction determined by 4D flow MRI and portography. Similar to our study, a previous study reported an increased amount of ascites after BRTO with improved hepatic synthetic function (30). In our cases, the small amount of ascites could have had little effect on the quality of 4D flow MR images. Nonetheless, it is important to consider the timing of image data acquisition since a large amount of ascites increases the radiofrequency pulse inhomogeneity, which causes image quality deterioration and measurement errors of flow velocity. Our findings demonstrated that 4D flow MRI is feasible for non-invasive assessment of dynamic changes in portal venous system circulation before and after BRTO. Future large-scale, multicenter studies are warranted to confirm its clinical relevance.

The present study has some limitations. First, the sample size was small. This pilot feasibility study could inform appropriate sample-size estimations for future large-scale studies. Second, we excluded some cases with invisible target vessels on 4D flow MRI, which further reduced the sample size. Blood vessel visualization can be improved using a gadolinium contrast agent or MRI techniques (31). Finally, we could not assess the clinical utility given our small sample size. Given the improved visualization of hepatopetal blood flow in the PV and Child-Pugh grade after BRTO, 4D flow MRI may allow detection of clinical liver function improvement. However, further studies are warranted.

In conclusion, we demonstrated that 4D flow MRI can be used to detect dynamic changes in the portal venous system circulation, including the quantitative blood flow rate and qualitative blood flow direction, before and after BRTO. However, a large-scale multicenter study is warranted to confirm its clinical relevance.

Footnotes

Acknowledgements

We are grateful to Taisuke Inoue, Mitsuaki Sato, Yuichiro Suzuki, Masaru Muraoka (First Department of Internal Medicine, University of Yamanashi), Hiroaki Watanabe (Department of Radiology, University of Yamanashi), and Kazuyuki Sato (Division of Radiology, University of Yamanashi Hospital) for their useful discussions.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was partially supported by JSPS KAKENHI (grand no. JP 21K15762).

ORCID iD

Masahiro Hamasaki

References

Khanna

Sarin

. Non-cirrhotic portal hypertension - diagnosis and management. J Hepatol 2014;60:421–441.

Sauer

Theilmann

Stremmel

, et al. Transjugular intrahepatic portosystemic stent shunt versus sclerotherapy plus propranolol for variceal rebleeding. Gastroenterology 1997;113:1623–1631.

Haska

. Improved patency of transjugular intrahepatic portosystemic shunts in humans: creation and revision with PTFE stent-grafts. Radiology 1999;213:759–766.

Hirota

Kobayashi

Maeda

, et al. Balloon-occluded retrograde transvenous obliteration for portal hypertension. Radiat Med 2006;24:315–320.

Hirota

Kobayashi

Kako

, et al. Balloon-occluded retrograde transvenous obliteration of varices: focusing on the portal hemodynamics and the recent techniques. Hepatol Int 2018;12:102–111.

Fukuda

Hirota

Matsumoto

, et al. Application of balloon-occluded retrograde transvenous obliteration to gastric varices complicating refractory ascites. Cardiovasc Intervent Radiol 2004;27:64–67.

Waguri

Osaki

Watanabe

. Balloon-occluded retrograde transvenous obliteration for treatment of gastric varices. World J Hepatol 2021;13:650–661.

Takuma

Nouso

Makino

, et al. Prophylactic balloon-occluded retrograde transvenous obliteration for gastric varices in compensated cirrhosis. Clin Gastroenterol Hepatol 2005;12:1245–1252.

Ninoi

Nishida

Kaminou

, et al. Balloon-occluded retrograde transvenous obliteration of gastric varices with gastrorenal shunt: long-term follow-up in 78 patients. AJR Am J Roentgenol 2005;184:1340–1346.

10.

Thabut

Moreau

Lebrec

. Noninvasive assessment of portal hypertension in patients with cirrhosis. Hepatology 2011;53:683–694.

11.

Owens

Bartolone

Warner

, et al. The inaccuracy of duplex ultrasonography in predicting patency of transjugular intrahepatic portosystemic shunts. Gastroenterology 1998;114:975–980.

12.

Chow

Soo

, et al. The measurement of liver blood flow: a review of experimental and clinical methods. J Surg Res 2003;112:1–11.

13.

Pieters

Miller

DeMeo

. Evaluation of the portal venous system: complementary roles of invasive and noninvasive imaging strategies. Radiographics 1997;17:879–895.

14.

Roldán-Alzate

Frydrychowicz

Niespodzany

, et al. In vivo validation of 4D flow MRI for assessing the hemodynamics of portal hypertension. J Magn Reson Imaging 2013;37:1100–1108.

15.

Burris

Hope

. 4D flow MRI applications for aortic disease. Magn Reson Imaging Clin N Am 2015;23:15–23.

16.

Dunås

Holmgren

Wåhlin

, et al. Accuracy of blood flow assessment in cerebral arteries with 4D flow MRI: evaluation with three segmentation methods. Magn Reson Imaging 2019;50:511–518.

17.

Roldán-Alzate

Francois

Wieben

, et al. Emerging applications of abdominal 4D flow MRI. AJR Am J Roentgenol 2016;207:58–66.

18.

Riedel

Lenz

Fischer

, et al. Abdominal applications of 4D flow MRI. Rofo 2021;193:388–398.

19.

Haarbye

Nielsen

Hansen

, et al. Four-dimensional flow MRI of abdominal veins: a systematic review. Diagnostics (Basel) 2021;11:767.

20.

Hyodo

Takehara

Mizuno

, et al. Time-resolved 3D cine phase-contrast magnetic resonance imaging (4D-flow MRI) can quantitatively assess portosystemic shunt severity and confirm normalization of portal flow after embolization of large portosystemic shunts. Hepatol Res 2021;51:343–349.

21.

Stankovic

Rössle

Euringer

, et al. Effect of TIPS placement on portal and splanchnic arterial blood flow in 4-dimensional flow MRI. Eur Radiol 2015;25:2634–2640.

22.

Bannas

Roldán-Alzate

Johnson

, et al. Longitudinal monitoring of hepatic blood flow before and after TIPS by using 4D-flow MR imaging. Radiology 2016;281:574–582.

23.

Roldán-Alzate

Frydrychowicz

Said

, et al. Impaired regulation of portal venous flow in response to a meal challenge as quantified by 4D flow MRI. Magn Reson Imaging 2015;42:1009–1017.

24.

Okuda

Yamada

Tanimoto

, et al. Three-dimensional cardiac cine imaging using the kat ARC acceleration: initial experience in clinical adult patients at 3 T. Magn Reason Imaging 2015;33:911–917.

25.

Tanihata

Minamiguchi

Sato

, et al. Changes in portal systemic pressure gradient after balloon-occluded retrograde transvenous obliteration of gastric varices and aggravation of esophageal varices. Cardiovasc Intervent Radiol 2009;32:1209–1216.

26.

Uehara

Akahoshi

Tomikawa

, et al. Prediction of improved liver function after balloon-occluded retrograde transvenous obliteration: relation to hepatic vein pressure gradient. J Gastroenterol Hepatol 2012;27:137–141.

27.

Lotz

Meier

Leppert

, et al. Cardiovascular flow measurement with phase-contrast MR imaging: basic facts and implementation. Radiographics 2002;22:651–671.

28.

Kim

Kweon

, et al. Multi-VENC acquisition of four-dimensional phase-contrast MRI to improve precision of velocity field measurement. Magn Reson Med 2016;75:1909–1919.

29.

Moersdorf

Treutlein

Kroeger

, et al. Precision, reproducibility and applicability of undersampled multi-venc 4D flow MRI sequence for the assessment of cardiac. Magn Reson Imaging 2019;61:73–82.

30.

Saad

Wagner

Al-Osaimi

, et al. The effect of balloon-occluded transvenous obliteration of gastric varices and gastrorenal shunts on the hepatic synthetic function: a comparison between child-pugh and model for end-stage liver disease scores. Vasc Endovascular Surg 2013;47:281–287.

31.

Frydrychowicz

Landgraf

Niespodzany

, et al. Four-dimensional velocity mapping of the hepatic and splanchnic vasculature with radial sampling at 3 tesla: a feasibility study in portal hypertension. J Magn Reson Imaging 2011;34:577–558.