ORIGINAL ARTICLE

Original Article

Modification of Trusler’s Formula for the

Pulmonary Artery Banding
Ghassan Baslaim, MD ∗
Division of Cardiothoracic Surgery, Department of Cardiovascular Diseases, King Faisal Specialist
Hospital and Research Center, MBC-J 16, P.O. Box 40047, Jeddah 21499, Saudi Arabia

Background: The major difficulty of pulmonary artery banding (PAB) is optimal intraoperative adjustment. In this
study, a target intraoperative shunt ratio (Qp /Qs ) was utilised to evaluate a fixed Trusler’s formula for the degree of PAB
of infants destined for either univentricular or biventricular surgical route.
Methods: Ten consecutive infants (median age, 1.62 months) undergoing PAB through median sternotomy were stud-
ied. A fixed Trusler’s formula (20 mm + 1 mm/kg body weight) was used to set the initial band size, and subsequent
intraoperative adjustment was based on target Qp /Qs (using oxymetric data). Suitable target Qp /Qs was set ≤0.5 for the
biventricular repair group (six patients) and ≤0.3 for the univentricular palliation group (four patients).
Results: In the biventricular group, the target mean Qp /Qs of 0.4 was achieved according to the fixed formula. However
the band size was narrower in the univentricular group by 2.25 mm to achieve the target mean Qp /Qs of 0.27 instead of a
higher level (2.2) with the fixed formula.
Conclusion: Using the intraoperative Qp /Qs calculation, the circumference of the band was in agreement with the fixed
formula of Trusler for the biventricular repair group, however a narrower band size is recommended for the univentricualr
palliation group.
(Heart, Lung and Circulation 2009;18:353–357)
© 2009 Australasian Society of Cardiac and Thoracic Surgeons and the Cardiac Society of Australia and New
Zealand. Published by Elsevier Inc. All rights reserved.
Keywords. Pulmonary blood flow; Congenital heart defects; Surgery

Introduction and contractility, mechanical ventilation, acid–base status,

haematocrit, and balance between systemic and pul-
T he purpose of any palliative surgical intervention is
to maintain balanced pulmonary-to-systemic blood
flow (shunt ratio) and to facilitate future surgical interven-
monary resistance. These variables are reflected on the
magnitude of pulmonary-to-systemic blood flow that can
be computed, as in the cardiac laboratory setup, by means
tion. However, determination of optimal band tightness is
of the oxymetric data [4].
often difficult, because even minor changes in the diame-
In view of the fact that a physiological method can
ter of the pulmonary artery have a large impact on the PBF
be utilised to estimate the pulmonary-to-systemic blood
and the gradient across the PAB. A simplified method for
flow, the intraoperative shunt ratio (Qp /Qs ) was imple-
banding the pulmonary artery was published known as
mented in this study to evaluate a fixed Trusler’s formula
Trusler’s formula [1,2]. The method is based on two prin-
for the degree of pulmonary artery banding of infants
ciples; the blood flow across the created pulmonary trunk
with increased PBF destined for either univentricular or
constriction, and the required blood flow according to the
biventricular surgical route. A target Qp /Qs was used to
infant weight.
determine the degree of pulmonary trunk constriction
The major difficulties of PAB are optimal intraoper-
intraoperatively provided acceptable systemic oxygen sat-
ative placement and adequate adjustment of the band.
uration is maintained.
Small changes in the diameter of the pulmonary artery
have a great effect on resistance and flow (Poiseuille’s
relation) [3]. Moreover, the effects of the banding on Materials and Methods
pulmonary artery pressure and flow are influenced by
clinical variables with mutual interference: heart rate Over a two-year period from April 2006, 10 consecutive
infants with complex congenital heart defects undergoing
pulmonary artery banding were studied by intraoperative
Received 4 January 2009; received in revised form 20 February shunt ratio (Qp /Qs ) calculation. The study was approved by
2009; accepted 20 February 2009; available online 1 May 2009 the institutional review board and informed consent was
∗Tel.: +966 2 667 7777x5234; fax: +966 2 663 9581. obtained from the parent of each patient. Age at banding
E-mail addresses: [email protected], [email protected] ranged from 0.62 to 5.9 months (median age, 1.62 months).

© 2009 Australasian Society of Cardiac and Thoracic Surgeons and the Cardiac Society of 1443-9506/04/$36.00
Australia and New Zealand. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.hlc.2009.02.003
 354 Baslaim Heart, Lung and Circulation
ORIGINAL ARTICLE

Modified pulmonary banding 2009;18:353–357

Table 1. Demography and Characteristics of Cases.

Future Route No. Age (month) Weight (kg) Diagnosis Indication for PAB Complication
Biventricle 1 1.64 1.8 VSD Prematurity
2 1.61 2.1 VSD Malnourishment
3 2.95 2.17 AVSD Malnourishment Sepsis
4 3.93 2.9 AVSD Malnourishment Pericardial effusion
5 5.9 4.4 mVSDs Repair complexity Pericardial effusion
6 4.92 4.6 cc-TGA, VSD Repair complexity
Univentricle 7 0.62 2.7 DILV PVP
8 0.62 2.98 TA, TGA PVP Prolonged intubation
9 0.92 3.9 DORV, hypo RV PVP
10 1.38 4.5 TA, TGA PVP Reintubation
PAB = pulmonary artery banding, VSD = ventricular septal defect, AVSD = atrioventricular septal defect, mVSDs = multiple ventricular septal defects,
cc-TGA, VSD = congenitally corrected transposition of great arteries and ventricular septal defect, DILV = double inlet left ventricle, TA, TGA = tricuspid
atresia and transposed great arteris, DORV, hypo RV = double outlet right ventricle and hypoplastic right ventricle, PVP = pulmonary vascular protection.

Their demographic data and clinical characteristics are and after each band adjustment. Suitable target Qp /Qs
shown in Table 1. Four of the 10 patients had univentricular was hypothetically set as less than or equal 0.5 for the
physiology, and banding was performed in preparation for biventricular repair patients and less than or equal 0.3 for
future cavopulmonary connection. The other six patients the univentricular palliation group. Acceptable intraoper-
were prepared for a delayed biventricular repair due to ative systemic oxygen saturation for future biventricular
malnourishment, prematurity or complexity of repair. repair and univentricular palliation was set as more
General anaesthesia was administered using Fentanyl than or equal 80% and 70%, respectively. Increased or
and Ketamine for induction and maintenance, and for decreased heart rate by 25% from baseline, signs of
muscle relaxation Pancuronium and Rocuronium were electrocardiographic ischaemia, hypotension, or hypoxia
used, to facilitate the intubation and the surgical proce- will lead to loosening of the band. The intraoperative
dure. Monitoring lines including central venous pressure transband gradient was not used as a guide for the band
and peripheral arterial pressure lines were inserted per- adjustment.
cutaneously, and distal pulmonary artery pressure line The procedures were performed through median ster-
was inserted directly through the median sternotomy. notomy and partial upper pericardiotomy. The associated
Packed red blood cells transfusion was utilised if neces- procedures such as ligation of the patent ductus arteriosus,
sary to stabilise the haemoglobin level (g/dl) unvaryingly and atrial septectomy (under inflow occlusion technique
throughout the procedure. Ventilatory settings were sta- through full anterior pericardiotomy) were performed, fol-
bilised all through the procedure with a fixed fraction of lowed by recording the baseline haemodynamics, heart
inspired oxygen (FiO2 ) for each case (in this study, FiO2 rate, haemoglobin level, different oxygen saturations and
was maintained ≤40%). Systemic blood pressure, systemic shunt ratio calculations. The tissue plane between ascend-
oxygen saturation, haemoglobin level and heart rate were ing aorta and pulmonary trunk was developed over a
monitored during the surgical procedure. limited area halfway between the sinotubular junction of
A fixed Trusler’s formula (20 mm + 1 mm for each kg the pulmonary trunk and origin of the right pulmonary
body weight) was used to set the initial band size and sub- artery. Pulmonary artery pressure was measured by plac-
sequent adjustment was based on individual physiologic ing a temporary catheter (Medtronic, Inc., Minneapolis,
variability of the calculated Qp /Qs provided stable haemo- MN) in the pulmonary trunk distal to the band position.
dynamics and acceptable systemic oxygen saturation. The The band material is a 3.2 mm wide strip of glue-
Qp /Qs was determined each time based on the oxymetric coated cotton tape (Ethicon; Johnson & Johnsosn Inc.,
saturations (O2 %) using the following formula: Somerville, NJ) with a radiopaque marker. Once circum-
ferential access of the pulmonary trunk was achieved, the
Qp aortic O2 % − superior vena cava O2 % band was placed around it and the band was tightened
=
Qs pulmonary venous O2 % − pulmonary arterial O2 % according to the fixed Trusler’s formula with subsequent
adjustments if necessary. The band was tightened by plac-
For each case, under homogeneous level of ing interrupted sutures of 5/0 non-absorbable polyester
haemoglobin and FiO2 , baseline and subsequent (TI-CRON; tyco, healthcare UK Ltd., Gosport, UK) in the
recording for each band adjustments were obtained band starting by the joining the premarked two points with
including; band size (mm), systemic arterial oxygen fine silk suture according to the fixed Trusler’s formula.
saturation (%), superior vena caval oxygen saturation Subsequent band tightening was done using the same 5/0
(%), and pulmonary artery (distal to the band) oxygen non-absorbable polyester suture material. In order to pre-
saturation (%). Pulmonary venous oxygen saturation was vent band migration, two sutures of 5/0 non-absorbable
assumed at a value of 99% due to technical difficulty in polyester were placed to anchor the band to the adventitia
obtaining the proper sample, and the baseline of systemic of pulmonary trunk.
oxygen saturation of all patients was above 92%. The Echocardiograms were obtained prior to hospital
resultant Qp /Qs were calculated at baseline (no band) dismissal and documented the transband gradient.
Heart, Lung and Circulation Baslaim 355

ORIGINAL ARTICLE
2009;18:353–357 Modified pulmonary banding

Echocardiography as well as cardiac catheterisation and band size in the univentricular palliation group by an
angiography were done before the consecutive surgery for average of 2.25 mm to achieve mean Qp /Qs and systemic
all univentricular palliation cases and for two out of four oxygen saturation: 0.27 and 77.8% instead of 2.2 and 83%,
in the biventricular repair group. respectively. No band loosening was required due to
the stability of heart rate, blood pressure and systemic
oxygen saturation. There was no demand for inotropic
Results use intraoperatively. Transthoracic Doppler echocardio-
The clinical profile of biventricular repair group and pal- graphy prior to discharge from the hospital showed a
liated univentricular group undergoing PAB is listed in transband peak gradient ranging from 55 to 90 mmHg
Table 1. The indications for PAB were three different rea- (mean gradient, 71.5 mm Hg) in all patients. The mean
sons in the biventricular repair group, but pulmonary gradient was 71.7 in the biventricular repair patients ver-
vascular protection was the only indication in the univen- sus 71.3 mmHg for the univentricular palliation group.
tricular group. Median duration of mechanical ventilation None of the patients required further adjustment of the
and intensive care unit stay was 2.5 and 4 days, respec- band.
tively. Complications were experienced in five cases Both groups were subjected to their consecutive opera-
during the postoperative period (Table 1). There were no tion except two patients of whom one (VSD) is delayed due
hospital deaths and no late deaths related to PAB. to socioeconomic reason (Table 3). The pulmonary band
Table 2 presents their operative results. Satisfactory slipping was not experienced in any patient, and the addi-
intraoperative shunt ratios and oxygen saturations, and tional pulmonary forward flow through the banded main
postoperative transband gradients were achieved. In the pulmonary artery was maintained at the time of Glenn
biventricular group, mean Qp /Qs (0.4) and systemic oxy- procedure in all univentricular patients.
gen saturation (89.2%) were achieved with the same The duration of PAB till the consecutive operation
estimated band size according to the fixed formula. How- was ranging between 3.5 and 21.5 months (mean of 12.4
ever the band size was narrower from the fixed formula months for the four patients in biventricular repair group,

Table 2. Band Sizes, Shunt Ratios, Oxygen Saturations, and Predischarge Transband Gradients.
Future Route No. Fixed PAB (mm) Final PAB (mm) I II III A (%) B (%) C (%) PABg1 (mmHg)
Biventricle 1 22 22 2.1 .42 .42 95 89 89 90
2 22 22 2.6 .37 .37 99 79 79 60
3 22 22 2.2 .48 .48 100 88 88 64
4 23 23 2.3 .36 .36 96 87 87 60
5 24 24 4.7 .33 .33 97 96 96 86
6 25 25 5.8 .43 .43 97 96 96 70
Univentricle 7 23 21 4.2 1.4 .21 94 85 79 55
8 23 21 5.8 1.7 .25 92 82 78 90
9 24 22 7 3.2 .32 95 85 76 70
10 25 22 6.6 2.4 .28 92 80 78 70
Fixed = Trusler’s formula, PAB = pulmonary artery banding, I = intraoperative prebanding Qp /Qs , II = intraoperative postbanding Qp /Qs with fixed
formula band size, III = intraoperative postbanding Qp /Qs with target shunt ratio and saturation, A = intraoperative prebanding systemic oxygen
saturations, B = intraoperative postbanding systemic oxygen saturations with fixed formula band size, C = intraoperative postbanding systemic oxygen
saturations with target shunt ratio and saturation, PABg1 = predischarge transband gradient.

Table 3. Band Sizes, Shunt Ratios, Oxygen Saturations, Transband Gradients, Band Duration and Consecutive Operations.
Future Route No. Fixed PAB Final PAB III IV C (%) D (%) PABg1 PABg2 PABd Consecutive
(mm) (mm) (mmHg) (mmHg) (month) Operation
Biventricle 1 22 22 .42 NCD 89 84 90 110 21.5 Repair
2 22 22 .37 – 79 – 60 – 16 Pending
3 22 22 .48 NCD 88 89 64 70 4 Repair
4 23 23 .36 – 87 – 60 – 2.2 Pending
5 24 24 .33 1.4 96 92 86 80 10 Repair
6 25 25 .43 1.6 96 88 70 95 14 Repair
Univentricle 7 23 21 .21 1.3 79 84 55 65 7 Glenn
8 23 21 .25 1.0 78 74 90 85 3.5 Glenn
9 24 22 .32 1.2 76 76 70 80 4.3 Glenn
10 25 22 .28 1.1 78 88 70 77 8 Glenn
Fixed = Trusler’s formula, PAB = pulmonary artery banding, Qp /Qs = shunt ratio, III = intraoperative postbanding Qp /Qs with target shunt ratio and
saturation, IV = preconsecutive operation Qp /Qs , C = intraoperative postbanding systemic oxygen saturations with target shunt ratio and saturation,
D = preconsecutive operation systemic oxygen saturations, PABg1 = predischarge transband gradient, PABg2 = preconsecutive operation transband
gradient, PABd = pulmonary artery banding duration, NCD = no cardiac catheterisation done.
 356 Baslaim Heart, Lung and Circulation
ORIGINAL ARTICLE

Modified pulmonary banding 2009;18:353–357

and 5.7 months for all patients in the univentricular the procedure for each case. (4) Increased or decreased
palliation group). Prior to the consecutive operation heart rate by 25% from baseline, signs of electrocar-
the documented mean Qp /Qs , oxygen saturation and diographic ischaemia, hypotension, or hypoxia was not
transband gradient were; 1.5, 88.3% and 88.8 mmHg, for experienced during PAB.
the biventricular repair group except two patients, and The pulmonary blood flow was expected to increase
1.2, 80.5% and 76.8 mmHg for the univentricular palliation with the drop in the pulmonary vascular resistance during
group (Table 3). growth of the children, and hence a tighter band was pre-
ferred for the univentricular palliation group to facilitate
a smooth cavopulmonary connection in the future. How-
Discussion
ever, for the purpose of the study design and in order to
During a PAB operation, the optimum degree of con- provide a better pulmonary vascular protection, the pre-
striction cannot always be achieved hence such surgical determined intraoperative appropriate Qp /Qs (less than
procedures continue to be associated with significant or equal; 0.5 for the biventricular repair patients and 0.3
postoperative morbidity and mortality. This has led to for the univentricular palliation group) and acceptable
the development of several adjustable PAB devices [6–8]. systemic oxygen saturation (more than or equal; 80% for
However, none of these devices have gained widespread future biventricular repair and 70% for univentricular pal-
clinical use since they are expensive special equipment, liation) were chosen hypothetically as safe conditions for
unreliable and lacking reproducibility of adjustments. the band adjustment. Moreover, since the PBF is more
Moreover, rare but serious complications have been pulmonary vascular dependant, transband gradient and
reported. These include erosion and necrosis of the pul- pulmonary artery pressure were not considered as indi-
monary artery, and pseudoaneurysm formation of the cators for the band adjustment.
pulmonary trunk [5,8]. In addition, the limitations were Although the transband gradient, in this study, was not
described included patient size and the anatomical rela- considered as a marker for the band adjustment, the pre-
tion of the great arteries. consecutive operation transband gradients correlated to
This pulmonary artery constriction is further compli- some degree with the prior to discharge from the hospi-
cated by the fact that during PAB, the patient is on tal measurements (Table 3). For both groups, in spite of
mechanical ventilation under general anaesthesia with the slightly higher intraoperative oxygen saturations and con-
chest open. Within the first few hours or days after oper- siderably lower shunt ratios; the preconsecutive operation
ation, there are significant alterations in heart rate and oxygen saturations were considered satisfactory (84–92%
contractility, partial pressure of oxygen and carbon diox- in the biventricular repair group, and 74–88% in the uni-
ide, acid–base status, haematocrit, and balance between ventricular palliation group), and the shunt ratios were at
systemic and pulmonary vascular resistance, and these acceptable level (average of 1.5 for the biventricular repair
factors compound the effect of the band. Nevertheless, group, and 1.2 for the univentricualr palliation group). This
in an effort to evaluate the degree of intraoperative pul- acceptable increase in the preconsecutive operation Qp /Qs
monary artery banding of infants with severe congestive for both groups is not the mere drop in the pulmonary
heart failure, a physiological method had been reported artery pressure and resistance over time, but the use of an
where serial arterial oxygen saturations determined the assumed value of 95% as the pulmonary venous oxygen
degree of pulmonary artery constriction [9]. Moreover, the saturation during the cardiac catheterisation led to a lesser
systemic oxygen saturation and transband gradient are denominator value in the oxymetric saturations formula.
evaluated as end points for band adjustment [3]. The fixed formula of Trusler’s (20 mm + 1 mm for each kg
In this study, the intraoperative shunt ratio (Qp /Qs ) body weight) was found to be suitable for the biventricular
was utilised as a physiological guidance to evaluate a repair group, whereas a narrower band was more appro-
fixed Trusler’s formula for the degree of pulmonary artery priate for the univentricular palliation group to facilitate
banding of infants with increased PBF. The shunt ratio future cavopulmonary connection. A modified calculation
is a reciprocal flow, and it is a reflection of the col- such as 18 mm + 1 mm for each kg body weight, should be
lective effect of heart rate, contractility, partial pressure considered for the purpose of pulmonary vascular protec-
of oxygen and carbon dioxide, and vascular resistance. tion in the univentricualr palliation cases.
For both groups; biventricular repair and univentricu- In this study, median sternotomy was preferred over
lar palliation, the Qp /Qs calculation was the endpoints lateral thoracotomy for several reasons; both lungs can
for band adjustment provided systemic oxygen saturation be completely ventilated throughout the entire procedure
was maintained at an acceptable level. The computation with the minimum ventilatory settings, cardiopulmonary
of Qp /Qs by means of the oxymetric data was used to bypass can instituted in unstable patients, associated
provide information on the magnitude of the pulmonary- procedures can be accomplished effectively, and better
to-systemic blood flow. accessibility for the pulmonary trunk. Accurate placement
Several factors were standardised and stabilised for the of the band so as not to encroach on the pulmonary artery
purpose of this study, such as: (1) median sternotomy branches or compromise the integrity of the pulmonary
approach, (2) the oxygen consumption value, which is valve was essential.
an imprecise measurement in paediatric patients, was In conclusion, although the number of cases enrolled
not implemented in the computation [4], (3) haemoglobin in this study was small and the pulmonary venous oxy-
level and FiO2 were maintained consistently throughout gen saturation was assumed, it is appropriate to utilise
Heart, Lung and Circulation Baslaim 357

ORIGINAL ARTICLE
2009;18:353–357 Modified pulmonary banding

such a physiological method that reflects multiple haemo- Karp RB, editors. Kirklin/Barratt-Boyes cardiac surgery. 3rd ed.
dynamic variables for band adjustment intraoperatively. Edinburgh, Scotland: Churchill Livingstone; 2003. p. 1113–75.
The band was adjusted according to acceptable target Chapter 27.
pulmonary-to-systemic blood flow primarily, provided [4] Vargo TA. Cardiac catheterization: hemodynamic measure-
acceptable systemic oxygen saturation is maintained. The ments. In: Garson Jr A, Bricker JT, Fisher D, Neish SR,
editors. The science and practice of pediatric cardiology. 2nd
circumference of the band was in agreement with the
ed. Baltimore, MD: William and Wilkins; 1998. p. 961–93.
fixed formula of Trusler for the biventricular repair group, Chapter 43.
however for the purpose of better pulmonary vascular [5] Corno AF, Bonnet D, Sekarski N, Sidi D, Vouhe P, Von
protection the band size was narrower by 2.25 mm for Segesser LK. Remote control of pulmonary blood flow: ini-
the univentricualr palliation group. Therefore, a modi- tial clinical experience. J Thorac Cardiovasc Surg 2003;126:
fied Trusler’s formula (18 mm + 1 mm for each kg body 1775–80.
weight) should be considered in the palliation of the uni- [6] Leeuwenburgh BPJ, Schoof PH, Steendijk P, Baan J, Mooi WJ,
ventricular patients when subjected to pulmonary artery Helbing WA. Chronic and adjustable pulmonary artery band-
banding. ing. J Thorac Cardiovasc Surg 2003;125:231–7.
[7] Schlensak C, Sarai K, Gildein HP, Beyersdorf F. Pulmonary
artery banding with a novel percutaneously, bidirectionally
References adjustable device. Eur J Cardiothorac Surg 1997;12:931–3.
[8] Michel-Behnke I, Akintuerk H, Valeske K, Thul J, Mueller M,
[1] Trusler GA, Mustard WT. A method of banding the pulmonary
Schranz D. Pseudoaneurysm of the pulmonary trunk after
artery for large isolated ventricular septal defect with and
placement of an adjustable pulmonary artery banding device
without transposition of the great arteries. Ann Thorac Surg
(FloWatch-PAB) in a patient with muscular ventricular septal
1972;13:351–5.
defect. J Thorac Cardiovasc Surg 2005;130:894–5.
[2] Albus RA, Trusler GA, Izukawa T, Williams W. Pul-
[9] Levitsky S, DuBrow IW, Hastreiter AR. Pulmonary artery
monary artery banding. J Thorac Cardiovasc Surg 1984;88:
banding in infants: a physiological intraoperative method of
645–53.
determining the effectiveness of the procedure. Ann Thorac
[3] Tricuspid atresia and management of single-ventricle physi-
Surg 1974;17:492–5.
ology.Kouchoukos NT, Blackstone EH, Doty DB, Hanley FL,