Role of Echocardiography in Detecting Portopulmonary

Hypertension in Liver Transplant Candidates

Cynthia L. Cotton,* Sanjay Gandhi,† Paul T. Vaitkus,† Malek G. Massad,‡
Enrico Benedetti,§ Robert G. Mrtek,㛳 and Thelma E. Wiley*

Portopulmonary hypertension (PPHTN) is a recognized estimation of PASP, the question of accuracy of this test
complication of end-stage liver disease that adversely in estimating pulmonary artery pressures still remains
affects the outcome of orthotopic liver transplantation
unanswered. In the setting of liver transplant patients,
(OLT). There are limited data on the role of Doppler
echocardiography in assessing pulmonary artery systolic there is limited data comparing the usefulness of echo-
pressure (PASP) in this population. The purpose of our cardiography in the assessment of PPHTN. The pur-
study was to examine the accuracy of Doppler echocardi- pose of our study was to examine the accuracy of Dopp-
ography in evaluating pulmonary artery pressures in liver ler echocardiography in evaluating pulmonary artery
transplant candidates. Clinical and demographic data pressures in liver transplant candidates.
were gathered retrospectively for 78 liver transplant can-
didates (48 men and 30 women, mean age 51 ⴞ 9.6 yr)
who had PASP determined both by right heart catheter-
ization (RHC) and echocardiography. Paired sample Methods
t-test was used to compare mean PASP by echocardiogra-
phy and RHC. Correlation of PASP between echocardi- We retrospectively reviewed the medical records of all patients
ography and RHC was determined using Pearson’s linear enrolled in the Liver Center who were undergoing evaluation
correlation. Positive and negative predictive values for for OLT at the University of Illinois at Chicago Medical
echocardiography for PASP > 50 mmHg are reported as Center between 1995 and 2002. One hundred eighty-seven
compared with RHC. The mean PASP by echocardiogra- patients were referred for cardiac workup before liver trans-
phy (43.2 ⴞ 12.3 mm Hg) was significantly higher than plantation. Of these, 15 patients did not have any workup
mean PASP by RHC (33.7 ⴞ 15.5 mm Hg; P < .001). done because of failure to show up for the appointment. The
Regarding PASP, there was a significant but weak corre- remaining 172 patients underwent echocardiograms; how-
lation between echocardiography and RHC (r ⴝ 0.46, ever, PASP could not be determined in 32 of these patients
P ⴝ .01). The positive and negative predictive values of
because of suboptimal windows. Twenty-two of the remain-
echocardiography for identifying clinically significant
pulmonary hypertension (PASP > 50 mm Hg) were ing 140 patients had a normal PASP (PASP ⬍ 30 mm Hg
37.5% and 91.9%, respectively. Echocardiography is a with no right ventricle [RV] enlargement or failure) and did
useful tool in estimating PASP in liver transplant candi- not undergo RHC. Thus, these patients were excluded from
dates. Patients with apparently elevated PASP by echocar- analysis. Forty patients did not undergo RHC despite meet-
diography should undergo invasive assessment by RHC ing criteria for abnormal PASP (PASP ⬎ 30 mm Hg or RV
before being excluded from liver transplant. (Liver failure or enlargement) because of the following reasons: 29
Transpl 2002;8:1051-1054.) patients were not listed because of other comorbidities and
were not considered transplant candidates, 2 patients died
while awaiting completion of the workup, 7 patients under-
P ortopulmonary hypertension (PPHTN) is a recog-
nized complication of end-stage liver disease that
adversely affects the outcome of orthotopic liver trans-
went liver transplant without preoperative RHC, and 2
patients were not listed secondary to severe PHTN by echo-
cardiogram. A total of 78 patients underwent both echocar-
plant (OLT). The prevalence of PPHTN varies from diogram and RHC and were included in the study. Patients
0.61% to 4%1,2 in patients with cirrhosis; however, the with PASP ⱖ 30 mm Hg, RV enlargement or dysfunction, or
prevalence in patients referred for OLT may be as high
as 8.5%.3,4 The wide variability in prevalence is partly
attributable to varying definitions of PPHTN. From the Departments of *Digestive and Liver Diseases, †Cardiol-
ogy, ‡Cardiothoracic and §Liver Transplant Surgery, and 㛳Medical Edu-
Direct measurement of pulmonary artery pressure cation, University of Illinois at Chicago, Chicago, IL.
by right-heart catheterization (RHC) remains the gold Address reprint requests to Thelma E. Wiley, MD, Section of Diges-
standard for detecting pulmonary hypertension. How- tive and Liver Diseases, Department of Medicine, 840 South Wood Street
ever, Doppler echocardiography using transtricuspid (M/C 716), Chicago, IL 60612. Telephone: (312) 996-6651; FAX:
valve gradient allows noninvasive estimation of pulmo- (312) 996-5103; E-mail: twiley@uic.edu
Copyright © 2002 by the American Association for the Study of
nary artery systolic pressure (PASP).5-9 Liver Diseases
Although small studies have been conducted previ- 1527-6465/02/0811-0010$35.00/0
ously regarding echocardiography and the noninvasive doi:10.1053/jlts.2002.35554

Liver Transplantation, Vol 8, No 11 (November), 2002: pp 1051-1054 1051

 15276473, 2002, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1053/jlts.2002.35554 by Sri Lanka National Access, Wiley Online Library on [25/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1052 Cotton et al

both underwent elective RHC. A PASP of ⱖ 50 mm Hg was

considered one of the criteria for categorizing patients in a
high-risk group for liver transplantation at our institution and
was therefore used as a cutoff to dichotomize the data. The
median time interval between echocardiogram and RHC was
21 days. There were 5 patients who underwent cardiac cath-
eterization for abnormal stress echocardiogram despite a nor-
mal PASP. Data was collected on age, race, etiology of liver
disease, and Child’s Class. Inclusion criteria were as follows:
patients with end-stage liver disease undergoing workup for
liver transplantation and those patients who underwent both
echocardiogram and RHC. Exclusion criteria included
patients who had only echocardiogram or RHC, patients with
suboptimal PASP by echocardiogram, or patients with evi-
dence of severe cardiac disease as evidenced by catheterization
procedure.
Echocardiography was performed with Acuson Sequoia
machines (Mountain View, CA). The studies were obtained
using multiple views including parasternal long and short Figure 1. Correlation between echocardiography and
axes, two-chamber apical and four-chamber apical views. Tri- RHC
cuspid regurgitation (TR) was characterized as mild, moder-
ate, or severe by qualitative assessment using Doppler color directly at end expiration with the catheter position in the
flow imaging. The PASP was estimated by adding the right main pulmonary artery.
atrial pressure to the transtricuspid gradient. The systolic
transtricuspid valve gradient was calculated from the peak TR Statistical Analysis
velocity using the modified Bernoulli equation 4v2, where v ⫽
Statistical comparisons were made with version 10 of the
highest continuous wave Doppler measurement of the TR
SPSS program (Prentice-Hall, Upper Saddle River, NJ). The
velocity in m/sec in the parasternal short axis and four apical
paired sample t-test was used to compare mean PASP by
chamber views. The PASP was determined by adding 10 or 14
echocardiography and RHC. The correlation between PASP
mm Hg to the transtricuspid gradient depending on the size
by echocardiography and RHC was determined using Pear-
of the inferior vena cava (IVC) as seen on subcostal views. A
son’s correlation coefficient. Positive and negative predictive
PASP of ⬎ 50 mm Hg was defined as significant in the
values of echocardiography are reported as compared with
assessment of patients undergoing liver transplantation. RHC
RHC.
was performed with standard techniques with a Swan Ganz
catheter and a fluid-filled hemodynamic system. When nec-
essary, patients were premedicated with oral Benadryl (25 mg) Results
and Diazepam (5 mg). Cardiac output was determined by
thermodilution, Fick method, or both. PASP was measured Baseline Demographics
Between 1995 and January 2002, 78 patients had PASP
measured by both echocardiography and RHC. There
Table 1. Etiology of Liver Disease were 48 (61.5%) men and 30 (38.5%) women. The
mean age was 51 ⫾ 9.6 years (range, 25-70 yr). The
Number mean Child-Pugh score was 10.5 ⫾ 2.3 (range, 6-15).
(Total ⫽ 78) Percentage The etiology of liver disease is summarized in Table 1.
ETOH 19 26 The mean PASP by echocardiography (43.2 ⫾ 12.3
HCV 18 24.7 mmHg) was significantly higher than that determined
HCV ⫹ ETOH 16 21.9 by RHC (33.7 ⫹ 15.5 mmHg) (P ⬍ .001). Sixteen of
Cryptogenic 7 9.6 78 (20.5%) had PASP ⱖ 50 mm Hg by echocardiog-
Auto-immune 4 5.5
raphy whereas only 9 out of 78 (11.8%) patients had
HBV 2 2.7
Other 7 9.6 PASP ⱖ 50 mm Hg by RHC. There was a significant
Unknown 5 6.4 but weak correlation between PASP by echocardiogra-
Abbreviations: ETOH, ethanol; HCV, hepatitis C virus; phy and RHC (r ⫽ 0.46, P ⫽ .01) (Fig. 1). The positive
HBV, hepatitis B virus. predictive value of echocardiography for identifying
clinically significant pulmonary hypertension (PASP ⬎
 15276473, 2002, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1053/jlts.2002.35554 by Sri Lanka National Access, Wiley Online Library on [25/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Echocardiography, PHTN, and Liver Transplant Candidates 1053

the clinical setting in which our patients’ data were

Table 2. PHTN Catheter Cross-Tabulation generated, there is a greater imperative to provide the
pressure estimate. Thus, our echocardiographers did
PHTN Catheter not have the luxury of selectively discarding studies that
Count No Yes Total were suboptimal. In our study, the right atrial pressure
PHTN-Echo was neither measured nor estimated and an empirical
No 57 5 62 number was added to derive the pulmonary artery pres-
Yes 10 6 16 sure. Because earlier studies used similar methods to
Total 67 11 78
estimate pulmonary artery pressure, this probably was
NOTE. As can be seen from the adjacent cross-tabs (Rows not a significant factor.6 One of the limitations of our
indicate values for echo and columns values for RHC), the
positive predictive value was 6/16 (37.5%), and the negative
study was that data were analyzed retrospectively. There
predictive value was 57/62 (91.9%). is some referral bias for performance of RHC because
most of the patients had an echocardiogram before
RHC. However, our study does include patients with
50 mm Hg) was 37.5%, whereas the negative predictive mild pulmonary HTN with PASP of 25 to 40 mm Hg
value was 91.9% (Table 2). There was no correlation of who underwent both echocardiography and RHC. We
pulmonary hypertension with age, gender, Pugh score, have calculated the positive and negative predictive val-
or liver disease. ues for predicting a PASP ⱖ 50 mm Hg, and the inclu-
sion of patients with a PASP of 25 to 50 mm Hg allows
for this calculation. To solve the issue of referral bias, a
Discussion prospective study with all patients undergoing RHC
Our study shows that echocardiography overestimates and echocardiogram would be needed, and this would
pulmonary artery pressures in patients with cirrhosis be difficult to justify in the absence of data to support
awaiting transplant as compared with RHC. Whether routine use of RHC for all patients. However, our study
we compared mean values for the two groups or we does include a significant number of patients with nor-
dichotomized the data according to a predefined crite- mal estimated pulmonary pressures who underwent
rion of PASP ⱖ 50 mm Hg, echocardiography resulted RHC and thus offers useful insights into the accuracy of
in higher pressure estimates than measured directly by echocardiogram in this subgroup of patients. Addition-
RHC. The utility of echocardiography in evaluating ally, the long time interval between echocardiography
these patients is best placed into context by considering and cardiac catheterization may result in variability.
the positive and negative predictive values. The positive Similar findings and limitations have been reported in a
and negative predictive values were calculated for sig- previous study.11
nificant pulmonary HTN is defined as PASP ⬎ 50mm In conclusion, echocardiography is a useful tool in
Hg, because it is used as a cutoff for considering patients estimating pulmonary artery pressures in liver trans-
at high risk for liver transplantation at our institution. plant candidates. In the case of normal or minimally
Because some patients with negative echocardiography elevated estimated pulmonary pressures, the echocar-
did not undergo RHC, the negative predictive value diographic results probably suffice. If the echocardio-
actually may be underestimated; therefore, our reported gram shows apparently elevated pulmonary pressures,
negative predictive value of 91.9 must be considered as invasive assessment is essential before rejecting the
the floor with the true negative predictive value being patient for a potential liver transplant.
even higher, reinforcing our conclusion that these
patients do not need a confirmatory RHC in most References
circumstances.
1. McDonnell PJ, Toye PA, Hutchins GM. Primary pulmonary
Our findings and those of Torregrosa et al10 differ hypertension and cirrhosis: Are they related? Am Rev Respir Dis
from initial studies validating the use of Doppler echo- 1983;127:437-441.
cardiography in assessing pulmonary pressures5-9 that 2. Castro M, Krowka MJ, Schroeder DR, Beck KC, Plevak DJ,
reported better correlations between the two modali- Rettke SR, et al. Frequency and clinical implications of increased
pulmonary artery pressures in liver transplant patients. Mayo
ties. One explanation for this difference is that in the
Clin Proc 1996;71:543-551.
original validation studies, a significant portion of 3. Ramsay MA, Simpson BR, Nguyen AT, Ramsay KJ, East C,
patients’ echocardiograms were deleted from analysis Klintmalm GB. Severe pulmonary hypertension in liver trans-
because of inadequate imaging of the regurgitant jet. In plant candidates. Liver Transpl Surg 1997;3:494-500.
 15276473, 2002, 11, Downloaded from https://aasldpubs.onlinelibrary.wiley.com/doi/10.1053/jlts.2002.35554 by Sri Lanka National Access, Wiley Online Library on [25/10/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
1054 Cotton et al

4. Auletta M, Oliviero U, Iasiuolo L, Scherillo G, Antoniello S. Morita T, et al. Noninvasive evaluation of pulmonary hyperten-
Pulmonary hypertension associated with liver cirrhosis: An echo- sion by a pulsed Doppler technique. Circulation 1983;68:302-
cardiographic study. Angiology 2000;51:1013-1020. 309.
5. Yock PG, Popp RL. Noninvasive estimation of right ventricular 9. Himelman RB, Stulbarg M, Kircher B, Lee E, Kee L, Dean NC,
systolic pressure by Doppler ultrasound in patients with tricus- et al. Noninvasive evaluation of pulmonary artery pressure dur-
pid regurgitation. Circulation 1984;70:657-662. ing exercise by saline-enhanced Doppler echocardiography in
6. Chan KL, Currie PJ, Seward JB, Hagler DJ, Mair D, Tajik chronic pulmonary disease. Circulation 1989;79:863-871.
AJ. Comparison of three Doppler ultrasound methods in the 10. Torregrosa M, Genesca J, Gonzalez A, Evangelista A, Mora A,
prediction of pulmonary artery pressure. J Am Coll Cardiol Margarit C, et al. Role of Doppler echocardiography in the
1987;9:549-554. assessment of portopulmonary hypertension in liver transplanta-
7. Currie PJ, Seward JB, Chan KL, Fyfe DA, Hagler DJ, Mair DD, tion candidates. Transplantation 2001;71:572-574
et al. Continuous wave Doppler determination of right ventric- 11. Kim WR, Krowka MJ, Plevak DJ, Lee J, Rettke SR, Frantz RP,
ular pressure: A simultaneous Doppler-catheterization study in et al. Accuracy of Doppler echocardiography in the assessment of
127 patients. J Am Coll Cardiol 1985;6:750-756. pulmonary hypertension in liver transplant candidates. Liver
8. Kitabatake A, Inoue M, Asao M, Masuyama T, Tanouchi J, Transpl 2000;6:453-458.