The Ross Procedure: Long-Term Clinical and

Echocardiographic Follow-Up
Nicholas T. Kouchoukos, MD, Paolo Masetti, MD, Nancy J. Nickerson, RN, MSN,
Catherine F. Castner, RN, William D. Shannon, PhD, and Victor G. Davila-Roman, MD
Division of Cardiovascular and Thoracic Surgery, Missouri Baptist Medical Center, Division of General Medical Sciences and
Biostatistics, Cardiovascular Imaging and Clinical Research Core Laboratory, Cardiovascular Division, Washington University
School of Medicine, St. Louis, Missouri

Background. Progressive dilatation of the pulmonary

autograft is the principal cause for reoperation following
the Ross procedure when the root replacement technique
is used. We examined the relation between enlargement
of the pulmonary autograft and the development and
progression of neo-aortic valve regurgitation, and the
long-term clinical follow-up, including the need for
reoperation, in patients followed for up to 13 years
postoperatively.
Methods. A Ross procedure was performed on 119
older children and young adults (mean age: 31 years old),
using the root replacement technique, between June 1989
and January 2002. Serial echocardiography studies were
obtained in 108 patients and analyzed blinded to clinical
data. The following variables were measured: diameter
of annulus, sinuses of Valsalva, and supravalvular ridge;
presence and severity of aortic regurgitation; and valve
thickening.
Results. The 30 day and late mortality rates were 1.7%
and 1.7% (2 patients each). Forty-one patients were followed for more than 5 years, 19 for more than 7 years, and
9 for more than 10 years. There was one thrombotic and
no endocarditis events. The 10-year actuarial survival
was 96%. Reoperation on the pulmonary autograft or the
pulmonary allograft was required in 12 patients. The

principal indication for operation on the pulmonary

autograft in 11 patients was neo-aortic valve regurgitation (7), aneurysm formation (3), and false aneurysm (1).
At 10 years, actuarial freedom from reoperation on the
pulmonary autograft was 75%. At last follow-up, 8 of 97
patients without reoperation on the autograft had moderate and none had severe regurgitation of the neo-aortic
valve. Independent predictors of progression of neoaortic valve regurgitation were time from operation,
dilatation of the supravalvular ridge, and neo-aortic
valve thickening (all p < 0.0002). Freedom from reoperation in the pulmonary allograft at 10 years was 86%.
Conclusions. Long-term follow-up of patients with the
Ross procedure using the root replacement technique
indicates excellent survival and low thromboembolic and
endocarditis risk. The main limitation is the need for
reoperation. The prevalence of severe neo-aortic valve
regurgitation is low, however there is a progressive
increase in regurgitation and in aortic root diameters.
Periodic follow-up with echocardiography is recommended because of the continuing risk of progressive
regurgitation of the neo-aortic valve and aneurysm formation.
(Ann Thorac Surg 2004;78:773 81)
2004 by The Society of Thoracic Surgeons

without regurgitation of the neo-aortic valve is a common

indication for reoperation [57]. Using serial echocardiographic studies, we assessed the function of the pulmonary autografts and examined the relation between enlargement of the pulmonary autografts and the
development and progression of neo-aortic valve regurgitation in patients undergoing the Ross procedure using
the root replacement technique who were followed for
up to 13 years postoperatively. We also evaluated the
long-term clinical outcomes, including the need for
reoperation.

eplacement of the aortic valve or aortic root with a

pulmonary autograft and replacement of the pulmonary valve with a pulmonary or aortic allograft (Ross
procedure) is a widely used technique for the treatment
of aortic valve disease in children and younger adults.
The procedure can be performed with low operative
mortality, and is associated with excellent hemodynamic
characteristics and an extremely low prevalence of infection and thromboembolism [13]. However, the need for
reoperation remains the principal limitation of the procedure [1, 4]. When the root replacement technique is
used, progressive dilatation of the autograft with or

Material and Methods

Accepted for publication Feb 10, 2004.
Presented at the Fiftieth Annual Meeting of the Southern Thoracic
Surgical Association, Bonita Springs, FL, Nov 1315, 2003.
Address reprint requests to Dr Kouchoukos, Cardiac, Thoracic and
Vascular Surgery, Inc, 3009 N Ballas Rd, St. Louis, MO 63131; e-mail:
ntkouch@aol.com.

2004 by The Society of Thoracic Surgeons

Published by Elsevier Inc

Patient Population
Between June 1989 and January 2002, 119 older children
and younger adults underwent a Ross procedure using
the root replacement technique. The patients ranged in
age from 4 to 56 years old (mean, 31 years old) and 83
0003-4975/04/$30.00
doi:10.1016/j.athoracsur.2004.02.033

CARDIOVASCULAR

ORIGINAL ARTICLES: CARDIOVASCULAR

CARDIOVASCULAR

774

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

(69%) were male. A bicuspid aortic valve, which was

present in 79 patients, was the most common indication
for operation. Fifteen of these patients also had aneurysms of the ascending aorta that were replaced with
polyester grafts. Twenty-seven patients had isolated aortic regurgitation, 9 had infective endocarditis (2 acute, 7
chronic), and 4 had failure of a previously inserted
porcine bioprosthesis or mechanical valve. Eleven of 119
patients had previously undergone balloon valvuloplasty
and 24 had undergone one or more open operations on
the aortic valve. Preoperatively, 20 of the patients were
New York Heart Association (NYHA) functional class I,
68 were class II, 28 were class III, and 3 were class IV. The
mean NYHA functional class was 2.1.

Ann Thorac Surg

2004;78:773 81

Table 1. Echocardiographic Assessment of Regurgitation in

the Pulmonary Autograft
Severity of
Regurgitation
None
Trivial
Mild
Moderate
Severe

Jet Width:
LVOT Diametera

Grade

0
0.2
0.2 0.29
0.3 0.39
0.4 0.49
0.5 0.59
0.6

0.0
0.5
1.0
1.5
2.0
2.5
3

The ratio of the width of the jet of regurgitation to the diameter of the
left ventricular outflow tract (LVOT) just below the level of the valve
annulus.

Preoperative Evaluation
In addition to echocardiographic and other noninvasive
studies, all adult patients underwent left-heart catheterization and coronary arteriography to assess the severity
of the aortic valve disease and left ventricular dysfunction, to detect the presence of coronary arterial occlusive
disease, and to determine the size and number of proximal septal perforating branches of the anterior descending coronary artery. Large conal branches of the right
coronary artery, which could be injured during removal
of the pulmonary root, were also identified.

Operative Procedure
Our technique for aortic root replacement with a pulmonary autograft and for reconstruction of the right ventricular outflow tract has been previously reported [3]. Intraoperative transesophageal two-dimensional and color
flow Doppler transesophageal echocardiography (TEE)
was performed in all but the youngest patients to measure the diameter of the aortic and the pulmonary annulus and to assess the competence of the pulmonary valve.
The diameter of the annulus of the pulmonary valve was
also measured after excision of the valve using a conical
sizer. Patients with marked dilatation of the aortic sinuses or with an annulus that measured more than 30
mm were not considered candidates for the procedure. In
the majority of patients, the diameter of the aortic annulus did not exceed the diameter of the pulmonary annulus by more than 2 to 3 mm. In 7 patients, plicating
sutures were placed in the aortic commissures to reduce
the diameter of the aortic annulus so that it approximated
the diameter of the autograft. A strip of polytetrafluoroethylene (PTFE) felt or autologous pericardium was used
to reinforce the suture line between the pulmonary
autograft and the aortic annulus in all of the older
children and adults. A strip of PTFE felt was also used to
reinforce the anastomosis between the pulmonary autograft and the ascending aorta in 65 of 119 patients. This
anastomosis was performed as close as possible to the
sinotubular junction of the autograft. The right ventricular outflow tract was replaced with a pulmonary allograft
in every patient. A graft with a diameter larger than the
diameter of the native pulmonary valve was used to
reduce the risk of development of pulmonary stenosis.

The allografts were not matched to the blood type of the

patient.
The mean duration of cardiopulmonary bypass was 192
minutes (range, 87 to 454 minutes). During the period of
myocardial ischemia, which averaged 149 minutes
(range, 54 to 222 minutes), the myocardium was protected by intermittent infusion of a cold blood cardioplegic solution directly into the coronary ostia or retrogradely into the coronary sinus. External cooling of the
heart was accomplished with a cooling jacket or topical
slush. After discontinuation of cardiopulmonary bypass,
the competence of the pulmonary autograft valve and the
pulmonary allograft valve in the right ventricular outflow
tract was assessed by TEE.

Associated Procedures
In addition to the 15 patients who had replacement of the
ascending aorta, 4 patients had replacement of the proximal aortic arch, 2 had repair of the mitral valve, and 1
each had coronary artery bypass grafting, closure of a
patent ductus arteriosus, repair of a coronary sinus-toright atrial fistula, left ventricular septal myectomy, and
repair of a foramen of Morgagni hernia.

Postoperative Evaluation and Management

Transthoracic M-mode, two-dimensional, color-flow, and
Doppler echocardiograms were obtained before discharge from the hospital and at 6 to 12 month intervals
thereafter. The severity of neo-aortic valve regurgitation
(AR) was measured using a modification of the method of
Perry and colleagues [8], in which the ratio of the width of
the jet of regurgitation to the diameter of the left ventricular outflow tract just below the level of the valve
annulus is determined. The regurgitation was graded as
follows: no AR, no diastolic color flow; trivial AR, ratio
more than 0 but less than 0.2; mild AR, ratio of 0.2 to 0.29;
mild to moderate AR, ratio of 0.3 to 0.39; moderate AR,
ratio of 0.4 to 0.49; moderate to severe AR, ratio of 0.5 to
0.59; and severe AR, ratio more than 0.6. A numerical
grade was assigned to these ranges (Table 1). Using the
method of Roman and coworkers [9], three dimensions of
the pulmonary root autograft at end-diastole were measured: the diameter of the annulus, the maximal diameter

of the sinuses of Valsalva, and the diameter at the

supravalvular ridge. Thickening of the valve leaflets was
qualitatively determined using a scale of 1 to 4 (1
normal, 4 severe valve thickening). The peak velocity
of flow across the pulmonary autograft and pulmonary
allograft valves was also measured, and gradients were
estimated using the modified Bernoulli equation. The
echocardiograms were interpreted by a single cardiologist who was blinded to the clinical status of the patients.
Serial echocardiography studies in 108 of 119 patients
were suitable for analysis. Four hundred and ninety-four
echocardiograms were analyzed (a mean of 4.5 studies
per patient). The severity of regurgitation of the neoaortic valve at the last visit was used for the comparative
analyses. Thirty-six patients had measurements of the
three dimensions of the autograft at baseline (within 6
months of operation) and 5 years postoperatively, 14 at
baseline and 7 years postoperatively, and 9 at baseline
and 10 years postoperatively.
No anticoagulants were administered to any of the
patients after operation. Use of platelet antiaggregating
agents in the follow-up period was left to the discretion of
the attending cardiologist. The mean duration of follow-up for the entire cohort was 4.4 years (range, 10
months to 13 years). The date of last inquiry was October
2002.

Statistical Analysis
Paired t-testing and pertinent mixed models analysis
were used to determine changes in diameter of the
neo-aortic root over time using SAS programs (Cary,
NC). Stepwise linear regression analysis was performed
to identify variables predictive of neo-aortic valve regurgitation. A p value less than 0.05 was considered statistically significant. Survival and event-free analyses were
performed using the Kaplan-Meier method.

Results
Mortality
The 30-day mortality was 1.7% (2 patients). Both patients
were women who developed severe right ventricular
dysfunction postoperatively. Temporary ventricular assist devices were inserted, but neither patient could be
weaned from the devices. Both patients had severe ventricular hypertrophy and small nondominant right coronary arteries, which could not be directly perfused, and
inadequate intraoperative myocardial protection was the
likely cause of the ventricular dysfunction. There were
two later deaths. One patient, a 12-year-old male with
severely depressed left ventricular function preoperatively, had persisting dysfunction postoperatively and
underwent cardiac transplantation. He died of a dysrhythmia 6 weeks following the Ross procedure. The
second patient, a 20-year-old male, developed a false
aneurysm at the proximal suture line of the pulmonary
autograft 2 months postoperatively. He died following
replacement of the autograft with an aortic allograft of

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

775

Fig 1. Survival following the Ross procedure. The number of patients at risk at 5, 7, and 10 years is shown in parentheses.
product limit estimation curve; E censored observations.

myocardial failure and a hemorrhagic diathesis. Although infection was suspected as the etiology of the
false aneurysm, no organisms were identified in blood
and tissue cultures. Survival for the entire cohort of
patients is illustrated in Figure 1. Forty-one patients were
followed for more than 5 years, 19 for more than 7 years,
and 9 for more than 10 years. The 10-year actuarial
survival was 96%.

Morbidity
Five patients (4.2%) required reoperation for bleeding.
Sixty-five patients (55%) did not require transfusion of
homologous blood products in the perioperative period.
Thirty-four of these patients had predeposited one or two
units of autologous blood which were transfused during
or after the operation. Among the 54 patients who required homologous blood, a mean of 3.8 U (range 1 to 36
U) of red cells was transfused.
Ventricular dysrhythmias requiring drug therapy occurred in 21 patients (18%). One patient required implantation of a permanent pacemaker. Seven patients had low
output syndrome. Two of these patients required insertion of ventricular assist devices and did not survive. One
of the remaining 5 patients had insertion of an intraaortic
balloon pump, and all survived. One patient developed a
temporary partial homonymous hemianopsia. The mean
and median postoperative lengths of stay in the hospital
were 6 and 7 days respectively (range, 3 to 47 days).
One patient with Libman-Sacks endocarditis and the
presence of antiphospholipid antibody, sustained a cerebral embolic event 9 months postoperatively and warfarin therapy was initiated at that time. No other patient
sustained a postoperative embolic event. The actuarial
freedom from thromboembolism at 10 years was 99%. No
patient developed endocarditis of the autograft or the
pulmonary allograft.

CARDIOVASCULAR

Ann Thorac Surg

2004;78:773 81

CARDIOVASCULAR

776

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

Ann Thorac Surg

2004;78:773 81

Table 2. Reoperations on the Autograft and the Pulmonary Allograft

Time
From
Initial
Operation
(months)

Age at
Initial
Operation
(years)

Gender

20

46

46

29

49

43

62

14

72

42

80

38

131

34

140

16

68

10

27

83

11

31

104

15

16

Autograft
only
1

Autograft and
allograft
9

Allograft only
12

Indication

False aneurysm of
autograft
AR, leaflet perforation and
prolapse
AR, aneurysm of ascending
aorta
AR, aneurysm of ascending
aorta, separation of
proximal suture line
Aneurysm of pulmonary
autograft

AR, aneurysm of
pulmonary autograft
AR, aneurysm of
pulmonary autograft
AR, aneurysm of
pulmonary autograft and
ascending aorta

AR, leaflet perforation and

prolapse, aneurysm of
pulmonary autograft, PR,
TR
AR,PS
AR, PS, and PR

PS

Operation

Aortic root allograft

AVR, replacement of
ascending aorta
Composite graft, CABG to
LAD
Composite graft,
replacement of
ascending aorta
Replacement of pulmonary
artery aneurysm,
resuspension of neoaortic valve
Composite graft, CABG to
R and LAD
Composite graft

Reoperation on the pulmonary autograft or the pulmonary allograft was required in 12 patients (Table 2). The
time interval between the initial procedure and the
reoperation ranged from 2 to 140 months. Eight patients
underwent reoperation on the pulmonary autograft, 3
patients on the autograft and the pulmonary allograft,
and 1 patient on the pulmonary allograft. The principal
indication for replacement of the 11 pulmonary autografts was regurgitation of the neo-aortic valve in 7,
aneurysm of the pulmonary arterial component of the
autograft in 3 (with or without associated regurgitation),
and false aneurysm in 1 patient. In 2 of 9 patients with
neo-aortic valve regurgitation, perforation of a leaflet,
likely related to injury at the time of implantation, was a
major cause of the regurgitation. In the remaining 7
patients, the regurgitation was associated with dilatation

Intraoperative
death
Survived
Survived
Survived

Survived

Survived
Survived

Replacement of pulmonary
artery and ascending
aorta, resuspension of
aortic valve

Survived

Composite graft,
replacement of
pulmonary allograft,
tricuspid annuloplasy
AVR, replacement of
pulmonary allograft
AVR, replacement of
pulmonary allograft

Survived

Replacement of pulmonary
allograft

Survived

AR aortic regurgitation;
AVR aortic valve replacement;
CABG coronary artery bypass grafting;
coronary artery;
PR pulmonary regurgitation;
PS pulmonary stenosis;
R right coronary artery;

Reoperation

Outcome

Survived
Survived

LAD left anterior descending

TR tricuspid regurgitation.

of the neo-aortic root and separation of the valve leaflets.

Only 3 of 11 procedures to replace or repair the pulmonary autograft were performed in the first 5 postoperative years. The remaining 8 procedures were performed
from 62 to 140 months following the initial operation.
Among the 4 patients who required replacement of the
pulmonary allograft, one, a 15-year-old female, had isolated stenosis of the truncal portion of the allograft. Of
the remaining 3 patients who also required replacement
of the pulmonary autograft, 1 patient had stenosis, 1 had
regurgitation, and 1 had both stenosis and regurgitation
of the allograft valve (Table 2).
At 5, 7, and 10 years postoperatively, freedom from
reoperation on the pulmonary autograft was 95%, 80%,
and 75% respectively (Fig 2), and freedom from reoperation on the pulmonary allograft was 99%, 92%, and 86%
respectively (Fig 3). The procedures performed at the

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

777

Fig 2. Freedom from reoperation on the pulmonary autograft. The

number of patients at risk at 5, 7, and 10 years is shown in parentheses. product limit estimation curve; E censored
observations.

Fig 4. Event-free survival for the 119 patients following the Ross
procedure. The number of patients at risk at 5, 7, and 10 years is
shown in parentheses. product limit estimation curve;
E censored observations.

time of reoperation are shown in Table 2. All but one of

the 12 patients survived reoperation and these 11 patients
were all alive at the date of last follow-up. Event-free
survival (freedom from death, reoperation, thromboembolism, and endocarditis) was 93% at 5 years, 78% at 7
years, and 73% at 10 years postoperatively (Fig 4).

ECHOCARDIOGRAPHIC MEASUREMENTS. The echocardiographic

Echocardiographic Studies
SEVERITY OF NEO-AORTIC VALVE REGURGITATION. The echocardiographically determined severity of aortic regurgitation at last follow-up for 97 patients with serial studies is
presented in Table 3. The 11 patients who required
reoperation are excluded. No patient had severe and 8
patients had moderate regurgitation.

dimensions of the pulmonary autograft at baseline and at

5, 7, and 10 years postoperatively, are listed in Table 4.
There was a consistent and substantial increase over time
in the mean dimensions at the sinus and supravalvular
ridge levels, but not at the annulus. The differences in the
dimensions at each time interval when compared to the
base line measurements for the same patients were
statistically significant at the sinus and supravalvular
ridge levels (p 0.02), but not at the annulus.
The relation between the severity of regurgitation of
the pulmonary autograft and the three dimensions of the
autograft as well as with the duration of follow-up is
illustrated in Figure 5. In a mixed models analysis, time
from operation, the diameter of the supravalvular ridge
and the thickening of the aortic valve were independent
predictors of the development of neo-aortic valve regurgitation (p 0.0001).
Continuous-wave Doppler echocardiography across
the pulmonary autograft at the most recent examination
demonstrated values for velocity that were within the
normal range. For the pulmonary allografts, the mean
Table 3. Severity of Aortic Regurgitation at Last Follow-Up

None
Trivial
Mild
Moderate

Fig 3. Freedom from reoperation on the pulmonary allograft. The

number of patients at risk at 5, 7, and 10 years is shown in parentheses. product limit estimation curve; E censored
observations.

Severe
Total
a

Numerical Grade

Number of Patients

0.0
0.5
1.0
1.5
2.0
2.5
3.0

22
46
14
7
8
0
0
97a

Eleven patients undergoing reoperation on the pulmonary autograft

excluded.

CARDIOVASCULAR

Ann Thorac Surg

2004;78:773 81

CARDIOVASCULAR

778

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

Ann Thorac Surg

2004;78:773 81

Table 4. Postoperative Echocardiographic Measurements of Autograft Diameters

Site
Annulus
Sinus
Ridge
a

Number

Baselinea
(cm)

Number

5 Years
(cm)

Number

7 Years
(cm)

Number

10 Years
(cm)

91
92
84

2.04 .31
3.08 .48
2.66 .40

36
36
35

2.15 .26
3.52 .48
2.92 .42

14
14
13

2.24 .28
3.82 .59
2.93 .43

9
9
9

2.23 .38
4.24 .63
3.22 .52

Within 6 months of operation.

computed peak systolic gradient across the allograft

valve for 36 patients followed for more than 5 years
(mean 7.7 years) increased from 7 5 mm Hg to 15
10 mm Hg (p 0.05). Four of 36 patients had gradients of
more than 25 mm Hg (27, 31, 36, and 52 mm Hg). Three of
these patients had no pulmonary regurgitation and one
had moderate regurgitation.

Comment
Our study confirms previous reports documenting the
safety of the Ross operation in selected children and

younger adults and a low prevalence of thromboembolic

complications and endocarditis. [2, 3, 10 13] Long-term
survival is excellent (Fig 1). [2, 4, 10, 12, 14] However, with
increasing follow-up, the need for reoperation remains as
the principal limitation of the procedure. This has been
observed with both the root replacement and the subcoronary techniques. [1, 4, 5, 14, 15] Although relatively
low reoperation rates have been reported in the first five
years postoperatively in many studies, our study and
others indicate that patients continue to develop substantial neo-aortic valve regurgitation after 5 years and require reoperation. [1, 2, 14] In our series, 8 of 11 reopera-

Fig 5. Boxplot analysis of severity of regurgitation of the neo-aortic valve plotted against the duration of follow-up and the three diameters of
the pulmonary autograft. The horizontal brackets represent the maximum and minimum observations and the darkly shaded areas encompass the 75th and 25th percentiles. The box width varies with the number of observations. In a mixed models analysis, length of follow-up
and diameter of the supravalvular ridge were independent predictors for development of neo-aortic valve regurgitation. (Top left) Follow-up.
(Top right) Annulus diameter. (Bottom left) Sinus diameter. (Bottom right) Ridge diameter.

tions on the pulmonary autograft were performed more

than 5 years postoperatively and as late as 11 years after
the initial procedure (Table 2).
Multiple causes for reoperation have been identified
with the root replacement technique. These include technical problems such as injury to the neo-aortic valve at
the time of implantation, geometric mismatch, false aneurysm, infection, immunologically mediated injury, and
structural changes in the autograft that predispose to
dilatation of the autograft with development of aortic
regurgitation and aneurysm formation. Progressive dilatation of the native aorta adjacent to the autograft may
also result in regurgitation of the neo-aortic valve because of dilatation at the site of anastomosis of the aorta
to the autograft. This latter change can occur with the
subcoronary, root inclusion, or root replacement techniques. In our study, which involved use of only the root
replacement technique, dilatation of the autograft with
progressive neo-aortic valve regurgitation was the most
common indication for reoperation.
When the root replacement technique is used, serial
echocardiographic studies have demonstrated progressive enlargement of the autograft over time [57, 10, 16,
17]. The dilatation has been most pronounced at the
sinus and sinotubular levels. In our series, no substantial
dilatation occurred at the annular level (Table 4). This
could be due to use of a band of PTFE or autologous
pericardium that reinforced the suture line. However,
dilatation at the level of the sinuses and the sinotubular
junction was significant and progressive, continuing for
up to 10 years postoperatively. Although this dilatation
was not consistently associated with the development or
progression of neo-aortic valve regurgitation, the diameter of the sinotubular ridge was an independent predictor for the development of AR.
Structural changes in the wall of the autograft and in
the valve leaflets are likely an important contributing
factor to development of dilatation and valve regurgitation when the root replacement technique is used [18
20]. In a histologic study of two of the autografts in our
series that were replaced at 52 and 80 months postoperatively, considerable transmural injury of the arterial wall
was evident, with loss of medial smooth muscle cells and
elastin and variable degrees of scarring, without inflammation or calcification [20]. Similar findings in two patients were reported by Takkenberg and colleagues [19].
Whether progressive neo-aortic valve regurgitation
and dilatation of the pulmonary autograft can be minimized or prevented by alternative techniques for implantation is not known with certainty. A follow-up study of
the pioneer series of Donald Ross reported by Chambers
and associates [1] demonstrated that neo-aortic valve
regurgitation was the most common indication for reoperation. Among 131 hospital survivors, 30 patients required late reoperations, and 28 of these procedures were
performed for severe regurgitation. Freedom from reoperation was similar for the 107 patients in whom the
subcoronary technique was used and for the 20 patients
in whom the root replacement technique was used. In a
more recent study, Sievers and coworkers [15] reported

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

779

their experience with the subcoronary technique in 228

patients and with the root inclusion technique in 17
patients who were followed for a mean of 29 months. The
reoperation rate on the autograft in the first 2 years was
comparable to that reported by Chambers and associates
[1].
Experience with the root inclusion technique is less
extensive. Although use of this technique may prevent
dilatation of the annulus and the neo-aortic sinuses, it
will not prevent progressive dilatation of the supravalvular aorta unless the ascending aorta is replaced or reinforced. Elkins and colleagues [2] reported a significantly
lower rate of freedom from reoperation and substantial
neo-aortic valve regurgitation among the 86 patients with
an intraaortic procedure (60 root inclusion, 26 subcoronary) than among the 242 patients in whom the root
replacement technique was used. Modification of the
root inclusion technique by fixation of the aortic annulus
and the sinotubular junction reported by Skillington and
coworkers [21] and David and associates [5] may improve
the outcomes. However, long-term follow-up information from these series is not available. Malposition and
distortion of the mobilized coronary arteries and other
technical problems may be limitations to the widespread
use of this technique [22].
Continued follow-up has confirmed earlier observations that moderate pressure gradients develop across
the pulmonary allografts in the right ventricular outflow
tract [3, 15, 23]. Patients with substantial gradients require reoperation [1, 3, 15, 23]. In our series, one patient
required reoperation for isolated stenosis of the pulmonary allograft, and two additional patients had the allograft replaced for stenosis in combination with replacement of the pulmonary autograft valve (Table 2).
Freedom from reoperation on the pulmonary allograft
was 86% at 10 years.
In conclusion, extended follow-up of young patients
having the Ross procedure using the root replacement
technique has demonstrated excellent survival with minimal risk for thromboembolism and endocarditis. Progressive dilatation of the autograft and regurgitation of
the neo-aortic valve that require reoperation remain the
principal limitations of the procedure. Reoperation for
these complications was performed more frequently in
the second 5-year interval of follow-up than in the first 5
years, and was required as late as 12 years postoperatively. Because of these continuing risks, periodic evaluation of all patients with echocardiography is
recommended.

Addendum
Since the date of last inquiry, an additional 7 patients
have required reoperation: 6 for neo-aortic valve regurgitation and dilatation of the pulmonary autograft (3 also
had replacement of the pulmonary allograft), and 1
patient for stenosis of the pulmonary allograft. The procedures to repair or replace the pulmonary autograft
were performed from 5.5 years to 11.5 years (mean 8.8
years) after the Ross procedure.

CARDIOVASCULAR

Ann Thorac Surg

2004;78:773 81

CARDIOVASCULAR

780

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

References
1. Chambers JC, Somerville J, Stone S, Ross DN. Pulmonary
autograft procedure for aortic valve disease. Circulation
1997;96:2206 14.
2. Elkins RC. The Ross operation: a 12-year experience. Ann
Thorac Surg 1999;68:S14 8.
3. Kouchoukos NT, Da vila-Roma n VG, Spray TL, Murphy SF,
Perillo JB. Replacement of the aortic root with a pulmonary
autograft in children and young adults with aortic-valve
disease. N Engl J Med 1994;330:16.
4. Elkins RC, Lane MM, McCue C. Pulmonary autograft reoperation: incidence and management. Ann Thorac Surg 1996;
62:450 5.
5. David TE, Omran A, Ivanov J, et al. Dilation of the pulmonary autograft after the Ross procedure. J Thorac Cardiovasc
Surg 2000;119:210 20.
6. Tantengco MVT, Humes RA, Clapp SK, et al. Aortic root
dilation after the Ross procedure. Am J Cardiol 1999;83:915
20.
7. Masetti P, Da vila-Roma n VA, Kouchoukos NT. Valvesparing procedure for the dilatation of the autologous pulmonary artery and ascending aorta after the Ross operation.
Ann Thorac Surg 2003;76:9156.
8. Perry GJ, Helmcke F, Nanda NC, Byard C, Soto B. Evaluation
of aortic insufficiency by doppler color flow mapping. J Am
Coll Cardiol 1987;9:9529.
9. Roman MJ, Devereux RB, Kramer-Fox R, OLoughlin J,
Spitzer M, Robins J. Two-dimensional echocardiographic
aortic root dimensions in normal children and adults. Am J
Cardiol 1989;64:50712.
10. Paparella D, David TE, Armstrong S, Ivanov J. Mid-term
results of the Ross procedure. J Card Surg 2001;16:338 43.
11. Prat A, Grandmougin D, Decoene C, et al. Aortic root
replacement with a pulmonary autograft in young adults:
medium-term results in 70 patients. Ann Thorac Surg 1998;
66:S148 52.
12. Oswalt JD, Dewan SJ, Mueller MC, Nelson S. Highlights of a
ten-year experience with the Ross procedure. Ann Thorac
Surg 2001;71:S3325.

Ann Thorac Surg

2004;78:773 81

13. Bo hm JO, Botha CA, Rein J-G, Roser D. Technical evolution

of the Ross operation: midterm results in 186 patients. Ann
Thorac Surg 2001;71:S340 3.
14. Laudito A, Brook MM, Suleman S, et al. The Ross procedure
in children and young adults: a word of caution. J Thorac
Cardiovasc Surg 2001;122:14753.
15. Sievers H-H, Dahmen G, Graf B, Stierle U, Ziegler A,
Schmidtke C. Midterm results of the Ross procedure preserving the patients aortic root. Circulation 2003;108[Suppl
II]:II-5560.
16. Savoye C, Auffray J-L, Hubert E, et al. Echocardiographic
follow-up after Ross procedure in 100 patients. Am J Cardiol
2000;85:854 7.
17. Simon-Kupilik N, Bialy J, Moidl R, et al. Dilatation of the
autograft root after the Ross operation. Eur J Cardiothorac
Surg 2002;21:470 73.
18. de Sa M, Moshkovitz Y, Butany J, David TE. Histologic
abnormalities of the ascending aorta and pulmonary trunk
in patients with bicuspid aortic valve disease: clinical relevance to the Ross procedure. J Thorac Cardiovasc Surg
1999;118:588 96.
19. Takkenberg JJM, Zondervan PE, van Herwerden LA. Progressive pulmonary autograft root dilatation and failure
after Ross procedure. Ann Thorac Surg 1999;67:5514.
20. Rabkin E, Aikawa M, Farber M, et al. Dynamic cell phenotypes mediate adaptive remodeling of clinical pulmonary
autograft valves in the aortic site. J Thorac Cardiovasc Surg
(in press).
21. Skillington PD, Fuller JA, Grigg LE, Yapanis AG, Porter GF.
Ross procedure. Inserting the autograft using a fully supported root replacement method; techniques and results.
J Heart Valve Dis 1999;8:593600.
22. Ross DN. Editorial. The subcoronary technique versus the
root technique for autograft surgery. J Heart Valve Dis
2002;12:5538.
23. Carr-White GS, Kilner PJ, Hon JKF, et al. Incidence, location,
pathology, and significance of pulmonary homograft stenosis after the Ross operation. Circulation 2001;104[Suppl I]:I16 20.

DISCUSSION
DR RONALD C. ELKINS (Oklahoma City, OK): Doctor Kouchoukos, I would like to compliment you on an outstanding
presentation and a very well thought out study. We have looked
carefully at our experience with the Ross operation and have
maintained long-term follow-up with 97% at two years, 84% of
these patients have had an echocardiogram within two years.
I have recently reviewed our patients, age 0 to 17 years, as I am
presenting a paper on these patients next week. In this group of
patients, we have an actuarial freedom from re-operation on the
autograft valve of 94% at 12 years, if the pulmonary valve was
normal at the time we did the operation. We have had excellent
hemodynamic results in these patients. We are concerned as you
are about the possibility of progressive dilatation.
We have done a multivariate analysis of the results in these
patients, and the only indicator that predicts re-operation on the
autograft valve is a primary lesion of aortic insufficiency (AI), not
a patient who had AI because he had a balloon valvuloplasty or
a previous aortic procedure for aortic stenosis but a primary
lesion of AI. In that group of patients, the Ross operation may
not be the appropriate procedure if you have another alternative, because at 12 years their freedom from re-operation is 64%,
a significant failure rate in these patients.
We have not completed the analysis similar to yours. I am
going back and relook at the change in pulmonary autograft root
size over time in our patients. I think it is probably not different

than yours. This is an outstanding paper and we appreciate your

presentation. Thank you.
DR KOUCHOUKOS: I think it is important if we are going to
continue to perform the Ross procedure, that we attempt to
identify the risk factors for dilatation and for progressive aortic
regurgitation. If the patients can be more appropriately selected,
the long-term results will be better.
DR ROSS M. UNGERLEIDER (Portland, OR): Doctor Kouchoukos, that is a wonderful follow-up and very helpful. We have had
a similar experience of about 160 Ross procedures on children
and adults between both Duke and now in Oregon. We have had
very limited amounts of autograft dilatation, but I was interested
in the fact that 66% of your patients had bicuspid aortic valves,
and we have been a little bit worried about the patients who
present with bicuspid aortic valves and dilated aortas to begin
with, and in those patients we have elected to replace the aortic
valve with an autograft and the ascending aorta with a Dacron
graft.
The one patient that had dilatation and failure of his autograft
in our series had a bicuspid aortic valve with an aneurysm of the
aorta at presentation and we didnt do that, and since then with
selective graft implantation for those patients we have havent

had a failure, but I dont know what that means; our follow-up
is only out to about 15 years.
I am curious if you had any thoughts about those types of
patients, that is, the patients who present with dilated aortas and
bicuspid aortic valves? Does that constitute a risk for eventual
autograft dilatation? Do you manage these patients with a
Dacron interposition graft in the aorta and if so, has that altered
the long- term outcome?
DR KOUCHOUKOS: Because of the limitations of time, I did
not present information on that issue. Fifteen of the patients in
our series had aneurysmal dilatation of their ascending aorta at
the time of the initial procedure and had the ascending aorta
replaced with a polyester graft. We have looked at those patients, and 2 of the 15 have required reoperation. Although one
would hope that there would be less of a problem with dilatation
of the autograft in this setting, I am not certain that replacing the
ascending aorta will totally avoid this problem.
DR THORALF SUNDT (Rochester, MN): Doctor Kouchoukos,
what advice do you have for us at the reoperation? Last week I
had occasion to do the sixth aortic valve operation on a 35-yearold woman. It can be a tough lie if they have had a previous Ross

KOUCHOUKOS ET AL
ROSS PROCEDURE: LONG-TERM FOLLOW-UP

781

and revisions thereof. You have said what you did with two of
your patients at redo. What advice do you have for us if we are
confronted with a patient with a previous Ross in terms of the
procedure that ought to be done and what advice would you
offer about the pulmonary homograft at that time?
DR KOUCHOUKOS: Well, the first piece of advice I would give
would be not to do a Ross operation when it is the fifth or sixth
procedure on the aortic valve. Clearly, reoperations are associated with increased risk. We have been fortunate that the 12
reoperations in our series resulted in only one death. This
patient was operated upon two months after the Ross operation
for a false aneurysm. We suspected infection but this was never
proven. Of the remaining 11 reoperations, two were valve
sparing procedures, one was an isolated replacement of the
pulmonary allograft, and the remainder were aortic root or
aortic valve replacements with or without replacement of the
pulmonary allograft.
One has to approach these reoperations with the same precautions as those taken for the other reoperations on the aortic
root or ascending aorta. The pulmonary allograft is also at risk
for injury, and it is important to know the location of a dilated
autograft or pulmonary allograft before dividing the sternum.

New Location for The Annals of Thoracic Surgery

Editorial Office
The Editorial Office of The Annals of Thoracic Surgery has
moved. As of June 18, 2004, our new address was:
The Annals of Thoracic Surgery Editorial Office
3440 Market St
Suite 306
Philadelphia, PA 19104-3325

2004 by The Society of Thoracic Surgeons

Published by Elsevier Inc

Ann Thorac Surg 2004;78:781 0003-4975/04/$30.00

CARDIOVASCULAR

Ann Thorac Surg

2004;78:773 81