Potts shunt in children with pulmonary arterial hypertension: institutional experience



There are many examples in the history of medicine of a well-known method receiving a ‘second life’: it is introduced for the treatment of a completely different disease. The Potts shunt, an anastomosis between the left pulmonary artery and the descending aorta, that has been used since 1946 as a palliation in patients with cyanotic congenital heart defects [1] is a good example of such a phenomenon. In 2004, Blanc et al. [2] performed a Potts shunt in 2 children (a 4-year old and 14-year old) with suprasystemic pulmonary arterial hypertension (PAH) and right ventricular failure. The authors suggested that ‘the advantages of the Potts procedure are the immediate decrease in the right ventricular afterload and the absence of cyanosis of the upper part of the body in the absence of intracardiac shunting’. The goal of performing the Potts shunt in this category of patients was to lower the suprasystemic PAH to a systemic level. The hypothetical rationale behind this approach was based on several studies showing that life expectancy in case of the Eisenmenger syndrome is better than that of idiopathic PAH. A combined retrospective and case–control study performed in 170 patients with Eisenmenger syndrome demonstrated that survival at 40, 50 and 60 years of age was 94%, 74% and 52%, respectively [3]. The median survival after diagnosis of idiopathic PAH recorded by the National Institutes of Health Registry was significantly shorter: 10 months for children and 2.8 years for adults [4]. However, the more recent REVEAL Registry, with 1979 outcomes analysed, failed to show better 4- and 7-year survival rates in patients with Eisenmenger syndrome [5].

Since 2004, the global experience with the Potts shunt in patients with PAH comprises about 50 interventions. However, the number of operations performed at 1 institution remains small [2, 6–9] (Table 1). The largest series that included 24 Potts shunts was described by Baruteau et al. [8]. However, these interventions were performed at 3 paediatric surgery centres. The aim of this publication was to present the results of the series of Potts shunts performed at the Bakoulev Center for Cardiovascular Surgery and to discuss the various aspects of this surgical approach.


This retrospective study was approved by the institutional review board of the Bakoulev Center for Cardiovascular Surgery with a waiver of consent. However, all of the parents and legal guardians of the patients provided general informed consent that allowed us to use the data received during the preoperative examination, the operative treatment and the follow-up visits for scientific analysis and publication.

During the years 2013–2016, 8 patients with PAH had the Potts shunt. Idiopathic PAH was diagnosed in 5 patients. One of them also had a patent ductus arteriosus (PDA) and 2 had a foramen ovale. Two patients had residual PAH after closure of the septal defect, and 1 had closure of the septal defect secondary to severe stenosis and hypoplasia of the pulmonary veins. In the last case, the diameter of the stenotic/hypoplastic pulmonary veins ranged from 1 to 2.5 mm.

Preoperative assessment included collection of historical data and a physical examination with determination of paediatric functional class [10] of arterial blood oxygen saturation (SaO2) at the upper and lower extremities. All patients underwent an echocardiographic examination. Catheterization of the right cardiac chambers was performed in all children except 1, who was in critical condition. Systemic pressure was measured in the radial artery. Left cardiac chamber catheterization was performed in only 1 patient with PDA. The 6-min walk distance (6MWD) test was only possible in 3 older children.

Six patients underwent the ‘traditional’ side-to-side Potts shunt between the left pulmonary artery and the descending aorta; in 1 case, a graft was used. One patient had a ‘functional’ Potts shunt by PDA stenting.

During their stay in the intensive care unit (ICU) after the operation, all patients underwent measurements of the SaO2 at the upper and lower extremities as well as echocardiography. At discharge, all patients who survived the time in the hospital were subjected to SaO2 measurements and echocardiography. The 6MWD was determined in the 3 patients who had performed this test preoperatively. Only the discharge data were reported, because they did not differ from those obtained in the ICU.

All 6 patients discharged from the clinic after having the Potts shunt were followed up for 2 to 32 (median 17) months. The follow-up assessment included a physical examination with determination of paediatric functional class, SaO2 measurements at the upper and lower extremities and echocardiography. The 6MWD test was performed in 3 patients, and cardiac catheterization/angiocardiography was performed in 1 patient.

Statistical analysis

Data were analysed using the SPSS 9.0 for Windows (SPSS, Chicago, IL, USA) and presented as mean ± standard deviation or median and range, where appropriate, or as absolute numbers and percentages. The parameters taken at multiple time points (before the operation, postoperatively and at follow-up) in patients who survived the hospital period were compared using analysis of variance for repeated measures followed by the t-test. P-values <0.05 were considered statistically significant.

Read more  Sedentary Lifestyle and Hypertension in a Periurban Area of Mbarara, South Western Uganda: A Population Based Cross Sectional Survey


The ages of the patients at the time of the operation ranged from 5 to 154 (median 13.5) months. Preoperatively, all of them were in the IIIb or IV paediatric functional class and had a weight deficit (Table 2). Six children reported syncopal episodes. All patients, except 1 with congenital stenosis and hypoplasia of the pulmonary veins, received 1- or 2-component specific PAH therapy but had stopped adequately responding to it. The aforementioned treatment did not include intravenous administration of epoprostenol, because this medication is not available in Russia. Patients were also given supportive therapy on a case-by-case basis, including diuretics, anticoagulants and oxygen. Preoperative cardiac catheterization revealed suprasystemic mean pulmonary arterial pressure in all cases (86.6 ± 11.9 mmHg). The pulmonary artery-to-systemic artery mean pressure ratio reached 1.39 ± 0.30. Echocardiographic examination showed suprasystemic right ventricular pressure, curvature of the ventricular septum and tricuspid regurgitation (Table 2). The PDA diagnosed in 1 patient was 3 mm in diameter. A small patent foramen ovale (2–4 mm in diameter) with a right-to-left shunt was noted in 2 patients. None of the patients, except the 1 with PDA, had a preoperative SaO2 gradient between the upper and lower extremities.

In 7 cases, the Potts shunt was performed via a left-sided thoracotomy. In 6 of them, it was ‘non-restrictive’, meaning that its diameter was not less than the diameter of the descending aorta. One patient with a pulmonary artery-to-systemic mean pressure ratio equal to 1.42 had a Potts shunt with a 7-mm Gore-Tex graft (the ratio of the graft-to-descending aorta cross-sectional area and the ratio of the graft cross-sectional area-to-body surface area were 0.49 and 7.6, respectively). In a patient with a PDA, a 7-mm stent was inserted via the left femoral artery. All patients survived the operation and were transferred to the ICU.

During the time in the hospital, the patients demonstrated a significant decrease in the lower limb SaO2 compared with preoperative levels and an SaO2 gradient between the upper and lower limbs (Table 2). All patients received inotropic support, anticoagulants and specific PAH therapy. The median duration of mechanical ventilation and of the stay in the ICU was 8 (range 2–27) and 24.5 (range 5–147) h, respectively. The hospital stay was uneventful in 6 patients. They demonstrated significant improvement in paediatric functional class at discharge (Table 2). Echocardiographic examination showed no signs of suprasystemic PAH. The ventricular septum was no longer inverted but flattening. The right ventricle-to-left ventricle end-diastolic diameter ratio significantly decreased compared with the preoperative values (Table 2). However, the decrease in the right ventricular pressure and tricuspid regurgitation did not reach statistical significance (P =0.11 and 0.22, respectively). Right-to-left shunting of blood from the right pulmonary to the ascending aorta via an anastomosis was present in all cases. Three patients who were capable of performing the 6MWD test increased the distance walked from 146, 128 and 132 m preoperatively to 442, 411 and 295 m at discharge, respectively.

The 2 patients who died during the hospital period had the highest preoperative pulmonary artery-to-aorta mean pressure ratio (1.79–1.86). The first patient was extubated after 9 h of mechanical ventilation and transferred to the floor from the ICU on postoperative Day 2. The lower limb SaO2 and the SaO2 gradient between the upper and lower limbs reached 65% and 25%, respectively. Despite continuing therapy with bosentan and sildenafil, the patient developed a severe pulmonary hypertensive crisis that led to death on postoperative Day 6. The second patient was extubated after 12 h of mechanical ventilation and transferred to the floor from the ICU on postoperative Day 2. However, the patient later experienced hypoxaemia of both the lower and the upper limbs (minimal SaO2 values were as low as 48% and 75%, respectively). This hypoxaemia, together with repeated episodes of pulmonary hypertensive crisis, was considered an indication for a Potts shunt narrowing on the fifth day after the initial operation. However, the patient died of heart failure during this repeat intervention.

At the follow-up visits, all children demonstrated a significant reduction in the weight deficit (Table 2). Maintaining the same specific PAH therapy, they did not experience any syncopal episodes or worsening of paediatric functional class compared with values obtained at discharge after the Potts shunt operation. Also, no significant changes occurred in the SaO2 gradient between the upper and the lower limbs or in any echocardiographic parameters. In all cases, there was right-to-left blood flow via the anastomosis. Two of the 3 patients who performed the 6MWD test had results that were slightly improved compared with those obtained at discharge (436 vs 411 and 311 vs 295 m, respectively). However, the third patient demonstrated some regression (335 vs 442 m) in the 2 years after the operation. Therefore, the patient underwent cardiac catheterization and angiocardiography that showed a functioning Potts shunt. The mean pulmonary arterial pressure was almost equal to the systemic (65 vs 63 mmHg) pressure, values that differed favourably from preoperative numbers (79 vs 59 mmHg). It should be mentioned that the results of the 6MWD test cannot be clearly interpreted, because it was performed in only 3 patients.

Read more  Stress Reduction Programs in Patients with Elevated Blood Pressure: A Systematic Review and Meta-analysis


Suprasystemic PAH in patients with intact cardiac septa leads to severe right chamber systolic overload and heart failure. Compression of the left cardiac chambers by enlarged right chambers and insufficient blood inflow to them because of high pulmonary resistance causes a critical decrease in systemic cardiac output. Suggesting a Potts shunt as a palliation for patients with PAH, Blanc et al. [2] emphasized that ‘the aim of this intervention was to decrease right ventricular afterload so as to improve right ventricular function and potentially prevent syncope and sudden death’. They also mentioned that ‘patients with severe right heart failure and markedly elevated pulmonary vascular resistance do not tolerate atrial septostomy, because massive right-to-left shunting may result in insufficient pulmonary blood flow and severe hypoxemia’. The advantages of a Potts shunt are as follows: (i) it is a constant ‘post-cardiac’ shunt, independent of atrial pressures, that provides decompression of the right cardiac chambers not only during the systolic phase but also during the diastole phase. (ii) Intervention decreases compression of the left ventricle by the right ventricle during systole. (iii) The procedure does not cause hypoxaemia of the upper part of the body that may have a negative impact on the brain and coronary function. (iv) There is no risk of paradoxical embolism of the brachiocephalic or coronary arteries. (v) The chance of shunt closure is low.

The Potts shunt is most often performed in patients with idiopathic suprasystemic drug-refractory PAH that is accompanied by severe right-sided heart failure and repeated syncopal episodes [8, 9, 11, 12]. In the majority of cases, including those in our series, the targeted population is infants or children. However, the Potts shunt has been used in adults as well [6]. Many of the Potts shunt candidates are already on a waiting list for a lung transplant. However, there is a discussion on using the Potts shunt at earlier stages, before the clinical condition reaches functional Class IV. Grady and Eghtesady [9] even consider this approach as an alternative to the infusion of a prostanoid. Hypothetically, it may improve the patient’s quality of life and prevent irreversible right ventricular failure. The Potts shunt has also been tested in experiments at the systemic or even subsystemic level of PAH. These studies demonstrated that a unidirectional valve implanted into a Potts anastomosis helps secure right-to-left shunting and to prevent its inversion, which could cause volume overload of the pulmonary circulation and an increase in pulmonary arterial pressure [13]. Experimental findings encouraged Baruteau et al. [8] to perform a ‘valved’ Potts shunt in a patient with a pulmonary-to-systemic artery mean pressure ratio of 0.82. The intervention was successful and resulted in an intermittent, unidirectional right-to-left shunt via an anastomosis. At the same time, the Potts shunt may not be indicated when suprasystemic PAH has already spontaneously decreased because of severely compromised right ventricular function. Grady and Eghtesady [9] reported a patient with a preoperative gradual decrease in pulmonary arterial pressure in whom the right ventricular function was not restored following a Potts shunt. The postoperative SaO2 upper/lower limb gradient was below 5%, and the patient died 3 months after the operation. A small series reported by Esch et al. [6] (Table 1) also illustrates the importance of preserved right ventricular function for the outcome of a Potts shunt: Only patients with normal or moderately impaired right ventricular systolic function survived the intervention. We believe that the Potts shunt can be performed regardless of age if the right ventricular systolic function is not severely altered. Patients with different levels of PAH can be considered possible candidates for the Potts shunt if they have the prominent aforementioned clinical manifestations and are in a low functional class despite intensive specific therapy. A ‘valved’ Potts shunt performed in patients with infrasystemic PAH may prevent deterioration of the clinical condition and syncopal episodes at exercise, when pulmonary arterial pressure and resistance are increasing. However, this intervention will probably not be beneficial if the pulmonary artery-to-aorta mean pressure ratio is below 0.7.

The 2 patients in our series who had the highest pulmonary artery-to-aorta mean pressure ratios (1.79 and 1.86) did not survive the hospital period and died of severe pulmonary hypertensive crisis and desaturation. This result raises the question of whether the ‘non-restrictive’ Potts shunt is beneficial to patients who have extremely high pulmonary arterial pressure. We hypothesized that, if the mean pulmonary arterial pressure significantly exceeds the systemic (by 20% or more) pressure, a ‘restrictive’ Potts shunt with a graft may be more appropriate. Positive experiences with a graft Potts shunt in 2 patients from the aforementioned population (the second patient underwent surgery during the period of the article’s editorial review) encouraged us to further evaluate this approach. Moreover, an optimal diameter of the anastomosis could be determined, taking into consideration the pulmonary artery-to-systemic mean pressure ratio and the cross-sectional area of the descending aorta or the body surface area. Our current belief is as follows: if the pulmonary artery-to-aorta mean pressure ratio is between 1.2 and 1.5, the ratio of the graft-to-descending aorta cross-sectional area should range from 0.5 to 0.4. If the pulmonary artery-to-aorta mean pressure ratio is above 1.5, the graft-to-descending aorta cross-sectional area ratio should be less than 0.4. However, data from our relatively small series are insufficient for coming up with valid conclusions and recommendations. Additional studies are needed. They should also be designed to determine the highest level of pulmonary arterial pressure when any type of Potts shunt is contraindicated.

Read more  Modifiable and non-modifiable risk factors for hypertension

There are few reports of ‘traditional’ or ‘functional’ Potts shunts in patients with suprasystemic PAH and existing or already repaired congenital heart defects. In 1 of these publications, PDA stenting in newborns or infants with different congenital heart defects and suprasystemic PAH not only lowered pulmonary arterial pressure to subsystemic levels but also made it possible to repair these defects at follow-up and perform a PDA closure [7]. We also performed a Potts shunt as the first stage of the operative treatment of a 5-month-old patient with severe stenosis and hypoplasia of the pulmonary veins, suprasystemic PAH and a poor clinical condition. Good functional results of the intervention have encouraged us to perform plastic reconstruction of pulmonary veins with less surgical risk in the near future. The Potts shunt was effectively combined with an atrial septostomy in patients with Shone’s complex, ‘borderline’ left ventricular dimensions and suprasystemic PAH [14]. The ‘hybrid’ intervention not only lowered pulmonary arterial pressure to the systemic level but also decreased left ventricular preload by creating a left-to-right shunt at an atrial level that led to a significant improvement of biventricular performance and regress of symptoms. Patients with a previously repaired ventricular septal defect or a complete atrioventricular canal but who experience suprasystemic PAH and right-sided heart failure may significantly benefit from a Potts shunt [15, 16]. We observed similar positive outcomes in our series when we performed the Potts shunt in 11- and 12-year-old patients with suprasystemic PAH after late ventricular septal or ventricular and atrial septal defect closure. In both cases, the Potts shunt resulted in lowering the pulmonary arterial pressure to the systemic level and significant improvement of the patients’ conditions.

In the majority of published series, an ‘open’ direct Potts shunt or anastomosis with a graft was performed. Grafts were mainly used in cases with limited mobility of the left pulmonary artery or a large distance between the left pulmonary artery and the descending aorta [9, 16]. However, different endovascular Potts shunt techniques were introduced as well [6, 8, 17–19]. Esch et al. [6] reported a transcatheter Potts shunt performed by aortic puncture opposite to the left pulmonary artery with implantation of a covered stent. However, 1 of the 4 operated patients died of a bleeding complication. Some authors believe that an endovascular Potts shunt can be safer if the surgeon uses radiofrequency perforation of the vessels instead of perforation by a trans-septal needle [18, 19]. Special magnetic catheters for the Potts shunt are being tested in different experiments [20].

Our findings are fully consistent with those of other researchers who showed that a Potts shunt significantly improves the functional status of patients and minimizes the chances of syncope; at the same time, hypoxaemia of the lower part of the body is moderate and well tolerated [2, 8, 9, 11, 12, 16, 21]. In line with Baruteau et al. [11], we have shown that a Potts shunt improves the altered ventricular dimensions (decreases right ventricle-to-left ventricle end-diastolic diameter ratio, per echocardiography). Probably due to the relatively small number of cases, we have demonstrated a tendency towards, but not a statistically significant postoperative decrease in, the right ventricular pressure, as was reported by the aforementioned authors. However, the fact that the ventricular septum is not inverted but flattening after the operation indicates that the pulmonary arterial pressure is no longer suprasystemic and that pressures in the ventricles are approximately equal.


Our work has multiple limitations. First, it is a retrospective study. Second, the number of cases may not be enough to prove a statistically significant difference when comparing some pre- and postoperative values. Finally, the follow-up period is too short to demonstrate long-term efficacy of the intervention. However, despite these limitations, we can conclude that a Potts shunt provides substantial benefits to patients with incurable suprasystemic PAH and may help them maintain a satisfactory condition while awaiting a lung transplant. In patients with concomitant congenital heart defects, it may provide an opportunity for the future repair of the defect. Further studies are needed to define more precisely the patient population for a Potts shunt as well as to determine the optimal diameter of the anastomosis.


The work was supported by funds from the Bakoulev Center for Cardiovascular Surgery, Moscow, Russia.

Conflict of interest: none declared.


Recommended For You

About the Author: Tung Chi