Portal Vein Pulsatility Ratio and
Heart Failure
Daniela Catalano, MD, Giuseppe Caruso, MD, Salvatore DiFazzio, MD, Giuseppe Carpinteri, MD,
Nunzio Scalisi, MD, Guglielmo M. Trovato, MD
Istituto di Medicina Interna e Terapia Medica, Università di Catania, Via Sant’Orsola 30, 95131 Catania, Italy
Received 5 March 1996; accepted 16 April 1997
ABSTRACT: Purpose. Heart diseases can alter liver
volume, morphology, and circulation. The Doppler
pulsatility of the portal vein and its pulsatility ratio
(PR) have been reported as being closely associated
with the right atrial pressure and with the New York
Heart Association (NYHA) class. We examined the
relationships between measurements of liver and
spleen dimensions and blood flow in portal and he-
patic veins, assessed noninvasively by Doppler so-
nography, and compared them with echocardiograph-
ic data.
Methods. The study group comprised 87 inpatients
with heart failure. The mean age was 64 ± 12 years.
Patients underwent duplex Doppler sonography of the
heart and portal and hepatic veins.
Results. Patients with more severe left ventricular
failure (NYHA class III–IV) showed more dilatation of
the left ventricle and atrium, reduced systolic function,
and reduced portal vein mean velocity compared with
patients with milder heart failure (NYHA class I–II); in
addition, the hepatic vein diameter was increased and
portal vein PR was reduced. Considering all patients,
significant positive correlations were found between
portal vein PR and left ventricular shortening fraction
(r = 0.34, p < 0.01) and ejection fraction (r = 0.38, p <
0.001). Significant negative correlations were found
between PR and hepatic vein diameter (r = −0.44, p <
0.001), right ventricle diameter (r = −0.38, p < 0.001),
left ventricular end-diastolic volume (r = −0.31, p <
0.01), and left atrium diameter (r = −0.33, p < 0.01).
Patients with hepatic vein dilatation had increased left
ventricular volumes, reduced systolic function indi-
ces, and portal vein alterations (increased diameter,
reduced mean velocity, and reduced PR). In patients
with an ejection fraction of no more than 50%, only PR
Correspondence to: G.M. Trovato
Presented in part at the Congress of Società Italiana di Ultra-
sonologia, 23–27 September 1995, Sorrento, Italy
© 1998 John Wiley & Sons, Inc. CCC 0091-2751/98/010027-05
VOL. 26, NO. 1, JANUARY 1998
was significantly reduced, while other sonographic
liver measurements were not significantly different.
Conclusions. The effects of cardiac failure on portal
blood flow, which declines progressively with wors-
ening cardiac function, is shown better by the pulsa-
tility pattern of the portal vein than by morphologic
caval and hepatic vein measurements. PR can be used
as a reliable adjunctive sign of heart failure. © 1998
John Wiley & Sons, Inc. J Clin Ultrasound 26:27–31,
1998.
Keywords: heart failure (congestive); portal flow; Dop-
pler abdominal ultrasonography; echocardiography
S everal cardiac diseases can alter liver volume,
morphology, and circulation.1–2 Severe con-
gestive heart failure causes liver enlargement, ca-
val and hepatic vein dilatation, and reduced or
absent caval collapsibility during inspiration.
Echocardiography is a reliable noninvasive
procedure for quantitatively assessing cardiac di-
mensions (atria, ventricles, and ventricular walls)
and systolic and diastolic function. Abdominal so-
nography can detect significant liver abnormali-
ties associated with congestive heart failure, not
secondary to pericardial, congenital, valvular, or
hypertrophic-restrictive cardiomyopathy and not
associated with primary liver disease. Recently,
the Doppler pulsatility of the portal vein and its
pulsatility ratio (PR 4 minimal peak velocity/
maximal peak velocity) were reported as being
closely associated with right atrial pressure and
New York Heart Association (NYHA) class.3–5
Doppler pulsatility evaluation, together with
measures of portal blood flow velocity, not only
provides information about the degree of periph-
eral venous dilatation and hypertension but also
gives some indication of alterations in portal
blood flow.
27
 CATALANO ET AL
The aim of the current study was to investigate
relationships between measurements of liver and
spleen dimensions and blood flow in portal and
hepatic veins, assessed noninvasively by Doppler
sonography, and compare them with heart func-
tion and dimensions assessed by echocardiogra-
phy.
PATIENTS AND METHODS
The study group comprised 87 inpatients (55
women, 32 men; mean age, 64 ± 12 years) who
had mild to severe heart failure. Disease severity
was determined according to the criteria of the
NYHA. Only patients with arterial blood pressure
of less than 160/95 mm Hg were included in the
study. Excluded were patients with diabetes mel-
litus; valvular, congenital, or acute ischemic
heart disease; chronic pericardial disease; fluid
effusion (pleural, pericardial, or abdominal) diag-
nosed clinically and by sonography; chronic liver
disease; anemia (hemoglobin < 12.5 g/dl); gastro-
intestinal bleeding; acute or chronic infection;
chronic renal disease (creatinine >1.5 mg/dl); lung
or pleural restrictive respiratory disease; or can-
cer. Severely obese patients were excluded as
well.
Duplex Doppler sonography of the abdomen
and heart was performed during a single session
on the patient’s second day of stay in the hospital.
Ansaldo Au530 ultrasound scanners (Genua,
Italy) equipped with 3.5–5.0-MHz sectorial and
convex-array probes were used. All echocardio-
graphic measurements were made following the
criteria of the American Society of Echocardiog-
raphy.6,7 Systolic and diastolic left ventricle sizes
and relative volumes (indexed for body surface
area), left atrium size, and right ventricle size
were measured in M and B modes. The ejection
fraction (EF), left ventricular shortening fraction,
stroke volume, cardiac output (CO), and several
other parameters were assessed as well. Echocar-
diography was performed by a single operator;
the intraobserver variability in the evaluation of
linear measurements was less than 3%, and vari-
ability in the evaluation of EF was 3.1 ± 1.2%.
The 2-dimensional echocardiographic studies
included long-axis and short-axis parasternal,
apical, and subcostal views.7 Practically, the left
ventricle was studied in 2 approximately orthogo-
nal apical views, including 1 4-chamber and 1
long-axis projection. The transducer was carefully
placed at the point of maximum impulse of the
cardiac apex. Wall motion was graded semiquan-
titatively on a 5-point scale: −1 4 hyperkinetic; 0
4 normokinetic; 1 4 hypokinetic; 2 4 akinetic; 3
28
4 dyskinetic. For analysis, the ventricle was con-
sidered to have 3 short-axis sections. The basal
section was from the junction of the interventricu-
lar septum and aortic root anteriorly and the mi-
tral valve annulus posteriorly to the level imme-
diately basal to the tip of the papillary muscles.
The midsection included the entire area of the
papillary muscles. The apical section was the area
from distal to the papillary muscles to the cardiac
apex. The basal and midventricular sections were
further divided into 5 segments each: septal, an-
terior, lateral, posterior, and inferior. The apex
was considered a single segment; thus, the ven-
tricle was divided into 11 segments. A wall motion
score8 was expressed as the sum of points in all 11
segments, theoretically ranging from −11 to 33.
Echocardiographic Doppler measurements were
not considered in data analysis because of the low
reproducibility of Doppler measurements in com-
parison with echocardiographic morphologic mea-
surements.9
After patients fasted overnight, we performed
Doppler sonography of the main portal vein proxi-
mal to its bifurcation and of the 3 main hepatic
veins via an intercostal or subcostal approach. As
a rule, the Doppler angle was maintained be-
tween 30° and 60°. Doppler samples were always
taken during apnea. The Doppler pulsatility of
the portal vein, the mean velocity of portal blood
flow, and the PR were measured using a sample
volume size 2⁄3 that of the vessel lumen. Liver and
spleen dimensions were assessed as well. The
mean portal vein PR in a group of 60 age- and
sex-comparable healthy subjects (ie, without liver
disease and with normal echocardiographic find-
ings) was 0.61 ± 0.09.
Statistical Analysis
Linear correlations between age, body mass
index, heart rate, and echocardiographic mea-
surements—systolic and diastolic volume, stroke
volume, CO, left atrium and right ventricle di-
mensions, left ventricular shortening fraction,
and EF—were assessed versus portal vein diam-
eter, mean blood flow, and PR; spleen maximal
length and hepatic vein diameters were also cor-
related with these measurements.
Patients were divided into 2 groups: NYHA
class I–II and NYHA class III–IV. Measurements
were compared in the 2 groups of patients by the
unpaired Student’s t-test. Patients were also com-
pared according to hepatic vein diameter [ø 9 mm
(normal) versus > 9 mm (dilated)] and EF (ø 50%
versus > 50%).
JOURNAL OF CLINICAL ULTRASOUND
 PORTAL VEIN PULSATILITY RATIO AND HEART FAILURE

RESULTS

Results for all patients and for patients in the

NYHA I–II (60 patients) and NYHA III–IV (27
patients) groups are summarized in Table 1.
Patients with more severe left ventricular fail-
ure (NYHA III–IV), compared with patients with
milder heart failure (NYHA I–II), showed greater
dilatation of the left ventricle and atrium, lower
systolic function, and lower portal vein mean
velocity; their hepatic vein diameters were in-
creased and their portal vein PRs were reduced
(Figure 1).
Considering all patients, significant positive
correlations were found between portal vein PR
and left ventricular shortening fraction (r 4 0.34, FIGURE 1. Hepatic vein diameter and pulsatility ratio in patients
p < 0.01) and between PR and EF (r 4 0.38, p < grouped according to severity of heart failure. Patients with more
severe heart failure (NYHA class III–IV) showed more dilated hepatic
0.001). Significant negative correlations were veins and reduced portal vein pulsatility ratios, ie, a less continuous
found between portal vein PR and hepatic vein flow in the portal vein.
diameter (r 4 −0.44, p < 0.001), PR and right
ventricle diameter (r 4 −0.38, p < 0.001), PR and
left ventricular end-diastolic volume (r 4 −0.31, p
other sonographic liver measurements did not
< 0.01), and PR and left atrium diameter (r 4
differ significantly between the low- and high-EF
−0.33; p < 0.004) (Figure 1).
groups (Table 3).
Compared with patients in whom the hepatic
vein diameters were normal, patients with he-
patic vein dilatation had increased left ventricle DISCUSSION
volumes, reduced systolic function indices, in-
creased portal vein diameters, reduced portal Congestive heart failure is a complex syndrome
vein mean velocity, and reduced portal vein PR with 2 prominent pathophysiologic features: (1)
(Table 2). inadequate or reduced cardiac output and (2) in-
Among patients with low EF (ø50%), only creased filling of venous compartments, ie, pe-
the portal vein PR was significantly reduced; ripheral, splanchnic, and pulmonary congestion.

TABLE 1
Cardiac and Hepatic Sonographic Measurements in Patients Grouped by NYHA Class

Mean ± Standard Deviation

NYHA I–II NYHA III–IV

Factor All (n = 60) (n = 27) p Value

Age (years) 64.1 ± 11.8 63.0 ± 11.8 66.6 ± 11.4 NS

BMI 27.2 ± 5.5 27.3 ± 5.7 27.2 ± 5.2 NS
LVEDV (ml/m2) 95.8 ± 35.2 82.8 ± 26.2 123.8 ± 36.0 <0.001
LVESV (ml/m2) 51.1 ± 32.5 37.7 ± 18.6 79.8 ± 37.6 <0.001
SV (ml/m2) 44.8 ± 12.4 45.0 ± 11.7 44.2 ± 14.0 NS
CO (l/min/m2) 3.4 ± 1.1 3.4 ± 1.0 3.5 ± 1.4 NS
LVMM (g/m2) 144.8 ± 39.6 136.2 ± 40.1 163.1 ± 32.2 <0.004
EF (%) 50.0 ± 13.9 55.8 ± 9.8 37.1 ± 13.1 <0.001
SF (%) 26.3 ± 8.6 29.8 ± 6.8 19.1 ± 7.3 <0.001
Asynergy score 2.1 ± 2.2 1.0 ± 1.4 4.3 ± 1.9 <0.001
LA (mm/m2) 24.6 ± 5.8 23.1 ± 6.0 27.5 ± 4.0 <0.001
PVD (mm) 11.0 ± 1.7 11.0 ± 1.7 11.0 ± 1.7 NS
PVmV (cm/min) 17.8 ± 5.3 18.5 ± 5.8 16.3 ± 3.6 <0.07
HV (mm) 9.4 ± 2.4 8.8 ± 2.1 10.7 ± 2.5 <0.001
PR 0.38 ± 0.23 0.45 ± 0.22 0.22 ± 0.16 <0.001

Abbreviations: NS, not significant; BMI, body mass index; LVEDV, left ventricular end-diastolic volume;
LVESV, left ventricular end-systolic volume; SV, stroke volume; CO, cardiac output; LVMM, left ventricular
myocardial mass; EF, ejection fraction; LA, left atrium diameter; PVD, portal vein diameter; PVmV, portal vein
mean velocity; HV, hepatic vein diameter; PR, pulsatility ratio.

VOL. 26, NO. 1, JANUARY 1998 29

 CATALANO ET AL

TABLE 2 Portal vein PR has been reported to be related

Cardiac and Hepatic Sonographic Measurements in to the right atrial pressure. It was supposed that
Patients Grouped by Hepatic Vein Diameter
the inferior vena cava and the hepatic veins, be-
Mean ± Standard Deviation cause they are in direct communication with the
HV > 9 mm HV ø 9 mm right atrium, dilate gradually; this blunts the ef-
Factor (n = 43) (n = 44) p Value fects of increased right atrial pressure on the por-
tal vein. When hepatic veins are maximally di-
Age (years) 64.3 ± 12.5 63.9 ± 11.2 NS
BMI 26.7 ± 4.5 27.7 ± 6.4 NS lated, hemodynamic events (ie, increased right
LVEDV (ml/m2) 106.6 ± 42.2 85.5 ± 22.9 <0.006 atrial pressure) can transduce the increased pul-
LVESV (ml/m2) 61.8 ± 40.6 40.9 ± 17.5 <0.003
satility to the portal vein through liver sinusoids,
SV (ml/m2) 44.8 ± 11.8 44.7 ± 13.0 NS
CO (l/min/m2) 3.4 ± 0.7 3.4 ± 1.4 NS reducing the portal vein PR.10 Our results showed
LVMM (g/m2) 151.5 ± 43.9 138.6 ± 34.6 NS a significant association between increased portal
EF (%) 46.3 ± 16.1 53.8 ± 10.1 <0.01
vein pulsatility (ie, reduced PR) and increased
SF (%) 24.1 ± 10.1 28.3 ± 6.4 <0.003
Asynergy score 3.3 ± 2.5 1.2 ± 1.3 <0.001 clinical severity of congestive heart failure (ie,
LA (mm/m2) 25.2 ± 5.0 24.0 ± 6.5 NS higher NYHA class) (Figure 1).
PVD (mm) 11.4 ± 1.7 10.6 ± 1.6 <0.03
Echocardiographic signs of congestive heart
PVmV (cm/min) 16.6 ± 4.2 19.1 ± 6.1 <0.03
HV (mm) 11.2 ± 1.9 7.6 ± 0.8 <0.001 failure—left atrial and ventricular dilatation, de-
PR 0.28 ± 0.22 0.48 ± 0.19 <0.001 creased EF, and hepatic vein dilatation—were
Abbreviations: see Table 1. also associated with a decrease in PR (Table 1).
Among patients with hepatic vein dilatation (Ta-
ble 2) echocardiographic and portal vein data both
These 2 components are usually referred to as indicated more severe congestive heart failure;
forward and backward cardiac failure. The cur- these findings included reduced portal vein PR,
rent sonographic approach to diagnosis of conges- reduced portal vein mean velocity, and increased
tive heart failure takes into account essentially portal vein diameter.
only backward cardiac failure, which is defined by Finally, when patients were grouped according
the degree of venous dilatation (caval and hepatic to the degree of impairment in left ventricular
veins) and by a decrease in or lack of venous col- systolic function, ie, according to the EF, the only
lapsibility on inspiration. Liver enlargement is discriminant liver sonographic measurement was
considered a consequence of this condition. Portal the portal vein PR (Table 3).
vein blood flow depends on several factors: CO, The portal vein PR provides information differ-
splanchnic flow, and intrahepatic microcircula- ent from that given by venous sonography, ie, de-
tion (sinusoids). It seems essentially related to gree of caval and hepatic vein dilatation and de-
CO, and alterations in CO must reflect the for- gree of venous collapsibility. These measures
ward heart failure component. show good agreement with clinical assessment of
heart failure by NYHA class. They also seem to be
signs of liver congestion. The effects of cardiac
TABLE 3 failure on portal blood flow, which declines pro-
Cardiac and Hepatic Sonographic Measurements in
Patients Grouped by Ejection Fraction
gressively with worsening cardiac function, can
be shown better by the pulsatility pattern of the
Mean ± Standard Deviation portal vein than by morphologic caval and hepatic
EF ù 50% EF < 50% vein measurements. Thus, the portal vein PR can
Factor (n = 50) (n = 37) p Value be used as a reliable adjunctive sign of heart fail-
Age (years) 61.7 ± 11.4 67.3 ± 11.3 <0.02 ure.
BMI 26.9 ± 5.3 28.2 ± 6.3 NS
LVEDV (ml/m2) 82.1 ± 25.4 112.9 ± 38.2 <0.001
LVESV (ml/m2) 33.9 ± 14.0 73.7 ± 35.8 <0.001 REFERENCES
SV (ml/m2) 48.1 ± 13.1 39.2 ± 9.6 <0.001
CO (l/min/m2) 3.6 ± 1.3 3.1 ± 0.8 <0.05
LVMM (g/m2) 137.0 ± 37.2 153.3 ± 40.2 <0.05 1. Bolondi L, Gaiani S, Barbara L: The portal venous
EF (%) 59.6 ± 7.0 37.4 ± 10.4 <0.001 system. In Cosgrove D, Meire H, Dewbury K, eds:
SF (%) 32.1 ± 5.4 10.8 ± 5.7 <0.001 Abdominal and General Ultrasound. Edinburgh,
Asynergy score 1.2 ± 1.8 3.7 ± 1.9 <0.001 Churchill Livingstone, 1993, p. 309.
LA (mm/m2) 24.1 ± 6.7 25.3 ± 4.2 NS
PVD (mm) 10.9 ± 1.6 11.1 ± 1.7 NS 2. Task Force on Heart Failure of the European So-
PVmV (cm/min) 18.8 ± 6.2 17.8 ± 5.0 NS ciety of Cardiology: Guidelines for the diagnosis of
HV (mm) 9.0 ± 2.1 9.8 ± 2.5 NS heart failure. Eur Heart J 1995;16:741.
PR 0.43 ± 0.24 0.30 ± 0.20 <0.006
3. Duerinckx AJ, Grant EG, Perrella RR, et al: The
Abbreviations: see Table 1. pulsatile portal vein in cases of congestive heart

30 JOURNAL OF CLINICAL ULTRASOUND

 PORTAL VEIN PULSATILITY RATIO AND HEART FAILURE
failure: correlation of duplex Doppler findings with
right atrial pressures. Radiology 1990;176:655.
4. Hosoki T, Arisawa J, Marukawa T, et al: Portal
blood flow in congestive heart failure: pulsed du-
plex sonographic findings. Radiology 1990;174:
733.
5. Wachsberg RH, Needleman L, Wilson DJ: Portal
vein pulsatility in normal and cirrhotic adults
without cardiac disease. J Clin Ultrasound 1995;
23:3.
6. Sahn DJ, DeMaria A, Kisslo J, et al: The Commit-
tee on M-mode Standardization of American Soci-
ety of Echocardiography: recommendation regard-
ing quantitation in M-mode echocardiography:
results of a survey on echocardiographic measure-
ments. Circulation 1978;58:1072.
7. Henry WL, DeMaria A, Gramiak K, et al: Report of
VOL. 26, NO. 1, JANUARY 1998
the American Society of Echocardiography Com-
mittee on Nomenclature and Standards in Two-
Dimensional Echocardiography. Circulation 1980;
62:212.
8. Gibson RS, Bishop HL, Stamm RB, et al: Value of
early two-dimensional echocardiography in pa-
tients with acute myocardial infarction. Am J Car-
diol 1982;49:1110.
9. Pozzoli M, Capomolla S, Cobelli F, et al: Reproduc-
ibility of Doppler indices of left ventricular systolic
and diastolic function in patients with severe
chronic heart failure. Eur Heart J 1995;16:194.
10. Catalano D, Caruso G, Carpinteri G, et al: The
liver in congestive heart failure: relationship be-
tween portal blood flow and echocardiographic as-
sessment. Journal of Heart Failure 1995;2:255.
31