An unusual cause of exercise dyspnoea and hypoxia after cardiac surgery: a case report

Abstract

Background

Dyspnoea is a cardinal clinical feature of post-cardiac surgery patients and represents one of the most significant postoperative complications, whether of cardiac or pulmonary origin.

Case summary

We report the case of a 52-year-old woman with mitral valve prolapse with severe regurgitation and a moderate-to-severe tricuspid regurgitation, who underwent mitral and tricuspid valvuloplasty using the Guiraudon approach. In the late postoperative phase, during cardiac rehabilitation, the patient experienced dyspnoea during low-intensity exercise associated with significant oxygen desaturation. A surgical technique-guided transthoracic echocardiography revealed a large atrial septal suture dehiscence (ASSD) with a significant bidirectional shunt, which was confirmed by transoesophageal echocardiography. An expert-performed cardiopulmonary exercise testing confirmed the pathophysiological link between ASSD and oxyhaemoglobin desaturation during effort. The patient subsequently underwent a second cardiac surgery to repair the ASSD, which led to the resolution of symptoms and a significant improvement in functional capacity, without further oxygen desaturation.

Discussion

This case underscores the importance of a multiparametric diagnostic approach and integrating expert-performed imaging and surgical technique-guided assessments in addressing the common challenge of post-cardiac surgery dyspnoea. The cardiac rehabilitation setting provides an ideal multidisciplinary environment for early detection and effective management of such complications.

Learning points.

• Unexplained dyspnoea after cardiac surgery should prompt targeted imaging.

• Knowledge of surgical techniques is essential to interpret diagnostic findings.

• CPET can reveal shunt-related hypoxemia not evident at rest.

Introduction

Dyspnoea is a cardinal clinical feature of post-cardiac surgery patients and often reflects the presence of significant post-surgical complications, whether cardiac or pulmonary in origin.^1–3 This clinical Case Report presents an unusual cause of dyspnoea and exertional oxygen desaturation following cardiac surgery. The case underscores the importance of a multidisciplinary and context-appropriate diagnostic and therapeutic setting for optimal patient management.

Summary figure

Summary figure about the clinical case Report.

Case presentation

A 52-year-old non-smoking woman with a history of mitral valve prolapse (MVP) presented with palpitations and underwent an unscheduled transthoracic echocardiogram (TTE), which revealed severe mitral regurgitation due to MVP, moderate-to-severe tricuspid regurgitation, normal pulmonary pressure and mild left ventricular (LV) dilatation with preserved ejection fraction (LVEF = 60%). No atrial septum defects or interatrial shunts were identified.

Following a coronary angiography that ruled out significant atherosclerotic disease, the patient underwent mitral (Physioring n°38 + neochord on P2) and tricuspid repair (Contour ring n° 32) via median sternotomy, with MV exposure using the Guiraudon technique.⁴ The atriotomies and the septum were closed with continuous 4-0 polypropylene monofilament sutures. Intra-operative transoesophageal echocardiography (TEE) confirmed a successful surgical outcome. The patient spent four uneventful days in the Cardiac Surgery Department before being transferred to the Cardiac Rehabilitation (CR) Unit.

At admission to CR, the patient was afebrile and haemodynamically stable with no signs of heart failure. The blood pressure values and the electrocardiogram were normal. Laboratory findings were unremarkable except for mild postoperative anaemia; inflammatory markers were within normal limits.

During the first rehabilitation session, the patient reported unusual exertional dyspnoea accompanied by mild resting hypoxaemia [arterial blood gas on room air, pH 7.46, oxygen partial pressure 58 mmHg, carbon-dioxide partial pressure 28 mmHg, oxygen saturation (SpO₂) 90%], with marked desaturation upon minimal exertion.

Chest radiograph and spirometry were normal.

A surgical-technique-guided TTE confirmed good valvular function and no significant regurgitation. LVEF was 50%, and no pericardial effusion was seen. However, a large atrial septal suture dehiscence (ASSD) with bidirectional shunt (left-to-right prevalent) was identified (diameter = 15 mm) with a pulmonary-to-systemic flow ratio (Qp/Qs) of 1.5 (Figure 1). Left atrial pressure surrogates, assessed using transmitral Doppler indices and the E/e′ ratio, did not indicate elevated filling pressures, supporting the absence of significant postoperative diastolic dysfunction or residual mitral valve obstruction. Right ventricular longitudinal systolic function was reduced, with a tricuspid annular plane systolic excursion of 8 mm, whereas radial contractility was preserved, as reflected by a right ventricular fractional area change of 51%. Right atrial pressure, estimated from inferior vena cava diameter and collapsibility, was within the normal range (∼5 mmHg), with no evidence of systemic venous congestion. Systolic pulmonary artery pressure could not be assessed due to the absence of tricuspid regurgitation following prior tricuspid annuloplasty.

Figure 1.

Two-dimensional TTE showing a large ASSD causing significant bidirectional (left-to-right prevalent) flow at colour Doppler. (A) apical four-chamber view, (B) parasternal short axis view at the level of the aortic valve, (C) subcostal four-chamber view.

Subsequent TEE confirmed excellent mitral and tricuspid valve repair and a large ASSD (diameter = 15 mm; aortic rim = 3 mm, posterior rim = 11 mm, superior rim = 12 mm, and inferior rim=16 mm). Colour Doppler demonstrated a predominantly left-to-right shunt, while contrast study with agitated saline revealed a right-to-left shunt (Figure 2) (see Supplementary material online, Videos S1 and S2 as Supplementary files). Specifically, colour-flow Doppler showed a broad, turbulent mosaic jet across the interatrial septum, predominantly directed from left to right. To optimize shunt visualization, a reduced Nyquist limit (35–40 cm/s) was applied, allowing clear detection of low-velocity interatrial flow. Shunt direction and velocity profile were further confirmed using pulsed-wave and continuous-wave spectral Doppler, which demonstrated a low-velocity, phasic flow pattern consistent with interatrial communication. Agitated saline contrast was injected via a peripheral antecubital vein using a standard three-way stopcock technique. Immediate opacification of the right atrium and right ventricle was observed, followed by the appearance of microbubbles in the left atrium within two cardiac cycles, consistent with direct intracardiac right-to-left shunting through the atrial septal suture dehiscence. No delayed bubble appearance beyond three cardiac cycles was observed, thereby excluding a significant intrapulmonary shunt mechanism.

Figure 2.

Two-dimensional TEE showing a large ASSD causing significant bidirectional shunt (left-to-right prevalent on colour Doppler and right-to-left prevalent after saline infusion mixed with air (>25 microbubbles)). (A) Mid-oesophageal bicaval view; (B) mid-oesophageal short-axis views with colour Doppler; and (C) mid-oesophageal short-axis view after saline infusion.

To further characterize the pathophysiological mechanisms responsible for both functional limitation and oxygen desaturation, an incremental cardiopulmonary exercise test (CPET) was performed. The test was prematurely terminated at 30 W due to severe dyspnoea and significant desaturation (SpO₂ dropped to 80%). CPET findings included a high ventilation-to-carbon-dioxide-output relationship (VE/VCO₂) slope and an abnormal reduction of end-tidal partial pressure of carbon dioxide (pETCO₂), consistent with multiple mechanisms, including a right-to-left intracardiac shunt during exercise (see Supplementary material online, Table S1—CPET1 as Supplementary file) (Figure 3).

Figure 3.

CPET panel plots before (left panels) and after (right panels) ASSD surgical repair. VO₂ = oxygen consumption; VE/VCO₂ = ventilation-to-carbon-dioxide-output relationship; pETCO₂ = end-tidal partial pressure of carbon dioxide; SatO₂ = oxygen saturation.

A multidisciplinary heart team, including the CR cardiologist, interventional cardiologist and cardiac surgeon, concluded that the shape and location of ASSD were unsuitable for percutaneous closure due to insufficient septal rim tissue, excessive defect size and iatrogenic aetiology.

A second cardiac surgery was performed via median sternotomy ten days after the initial procedure. The dehiscence, caused by suture thread rupture, was repaired. The postoperative phase was uneventful. Follow-up TTE demonstrated competent mitral and tricuspid repairs with no residual shunt, including after bubble testing (Figure 4) (see Supplementary material online, video  S3 and S4 as Supplementary files). Repeat CPET showed marked improvement in functional capacity without desaturation or abnormal ventilatory responses (see Supplementary material online, Table S1—CPET2 as Supplementary file) (Figure 3).

Figure 4.

Two-dimensional TTE showing the absence of interatrial shunt after surgical correction of ASSD. (A) Apical four-chamber view, (B) parasternal short-axis view at the level of aortic valve, and (C) subcostal four-chamber view.

The patient was finally discharged in stable condition and reported subjective well-being. At discharge, the patient had no medication left, except for anticoagulation therapy, sinus rhythm 80 bpm, and no signs of pulmonary congestion or pleural effusion on chest X-ray.

Discussion

The postoperative phase of patients undergoing cardiac surgery remains crucial for the overall success of the interventions, for the management of symptoms and complications and for the optimization of cardiovascular risk factor control and therapy. CR centres are recommended as increasingly recognized optimal settings for comprehensive post-surgical care. They enable integrated clinical and instrumental assessment, supervised exercise training and evidence-based secondary prevention strategies.^5,6 Among the symptoms in this phase, dyspnoea is one of the most frequent, with multiple pathophysiological bases (see Supplementary material online, Table S1 as Supplementary file). Clinical, laboratory, and echocardiographic evaluations are essential for diagnosis, but performing a CPET often proves to be ultimately decisive, being a valuable tool for assessing the aetiology and severity of dyspnoea and exercise intolerance.⁷

Following cardiac surgery, anaemia is common and often moderate to severe according to the World Health Organization classification. Nevertheless, peripheral oxygen saturation does not reliably reflect anaemia-related tissue hypoxia, as compensatory mechanisms preserve arterial saturation except in severe anaemia. Moreover, haemoglobin is a key determinant of oxygen delivery at rest and during exercise and, in the present case, a mild haemoglobin reduction of 4 g/dl would theoretically account for a decrease in peak VO₂ of ∼ 400 ml/min, assuming a peak cardiac output of 10 L/min. However, the observed CPET abnormalities cannot be explained by haemoglobin deficiency alone. This interpretation is further supported by the significant differences between CPET2 and CPET1 (see Supplementary material online, Table S1), despite comparable haemoglobin levels at both evaluations.

Adequate surgical exposure is crucial for successful mitral valve replacement or repair. The standard and most widely preferred approach involves a left atrial incision through the right interatrial sulcus, accessed via median sternotomy or mini-thoracotomy. Alternative approaches have been developed for complex cases, particularly when dealing with a small left atrium or in re-do procedures. The extended vertical trans-atrial septal approach, elegantly described in 1991 by Gérard Guiraudon,⁴ offers significant advantages and can be employed routinely. This technique provides several key benefits: Preservation of valve geometry, atrial function and superior anatomy and low complication profile. This approach has proven particularly valuable in cases requiring simultaneous mitral and tricuspid interventions, offering superior visualization without compromising cardiac function. Atriotomies are typically closed using 3-0 monofilament running sutures, beginning at the lower end and proceeding upward. In Guiraudon’s report of 34 consecutive patients, no bleeding or postoperative suture dehiscence was observed. While iatrogenic atrial septal defects are well recognized following transseptal catheter procedures, to our knowledge, no cases have been previously described after a Guiraudon procedure. This case therefore broadens the known spectrum of potential postoperative complications associated with the trans-atrial septal approach, underscoring the need for systematic echocardiographic assessment of the interatrial septum after such surgeries. Awareness of this possibility is essential for appropriately evaluating unexplained post-surgical symptoms.

Disruption or dehiscence of the interatrial septum following a trans-atrial approach can result in a direct communication between the atria. This complication can be identified through targeted, surgery-technique-guided TTE and confirmed by TEE, which provides detailed characterization of the defect, including size, shunt direction, right heart volume and function, and pulmonary pressures.^8–11 Saline contrast studies may be necessary to detect right-to-left shunting. TTE-derived pulmonary-to-systemic flow ratio (Qp/Qs) remains a validated tool to quantify haemodynamically significant shunting.⁸ In this case, the integration of surgical knowledge with expert echocardiographic assessment was the key to diagnosing ASSD, an uncommon but clinically significant postoperative complication.

Dynamic shunt reversal during exercise, often observed in conditions such as patent foramen ovale or atrial septal defect, refers to a shift from the usual left-to-right shunt (oxygenated blood entering the pulmonary circulation) to a right-to-left shunt (deoxygenated blood entering the systemic circulation). This occurs due to exercise- induced increases in right-sided cardiac pressures, commonly driven by transient pulmonary hypertension or altered ventricular compliance, which overwhelm the left heart’s filling capacity and force blood across the septal defect in the opposite direction. Thus, the pathophysiological mechanisms include elevated right heart pressure during exercise and ventricular compliance mismatch, leading to acute shunt reversal. If this reversal becomes chronic, it may contribute to the development of pulmonary hypertension, systemic hypoxaemia and cyanosis. In this specific clinical case, CPET further contributed to the diagnosis, being able to identify, among the various causes of dyspnoea, the cardiopulmonary consequences of an intracardiac shunting.^12–15 Evidence from patients with congenital heart disease or pulmonary arterial hypertension with a patent foramen ovale shows that exercise-induced changes in arterial oxygen saturation and end-tidal CO₂ (pETCO₂) can reveal such shunting. As pulmonary vascular resistance rises during exercise, right-to-left flow increases, potentially reversing a left-to-right shunt. In healthy subjects, oxygen uptake and carbon dioxide output rise proportionally, with stable saturation. In contrast, patients with right-to-left shunting exhibit falling oxygen saturation, reduced pETCO₂, and increased pETO₂ due to compensatory hyperventilation, even without resting pulmonary hypertension. During exertion, reversal of a left-to-right shunt may occur, causing deoxygenated, CO₂-rich blood to enter systemic circulation. This leads to reduced SpO₂ and increased VCO₂. Chemoreceptor-driven hyperventilation in response to elevated CO₂ levels results in an abnormal decline in pETCO₂ during exercise. In our patient, CPET demonstrated early exercise termination due to significant SpO₂ reduction, a VE/VCO₂ slope at the upper normal limit and a pathological drop in pETCO₂, suggesting a right-to-left shunt across the ASSD, even if CPET variables are physiologically non-specific and may reflect multiple mechanisms, including ventilation-perfusion mismatch, altered pulmonary haemodynamics, or right ventricular functional impairment. Following surgical repair, repeat CPET showed marked improvement in exercise capacity and normalization of parameters of ventilation with no oxygen desaturation (Figure 3).

In conclusion, dyspnoea and oxygen desaturation at rest and during exercise are frequent after cardiac surgery. A multidisciplinary and multiparametric approach is critical for accurate diagnosis and effective management. This case highlights the essential role of expert-performed, surgery-technique-guided imaging and CPET within the CR setting, which proved pivotal for identifying and resolving an uncommon but impactful postoperative complication.

Lead author biography

Dr. Silvia Di Marco graduated at the University of Milan. Her main areas of interest are echocardiography, pulmonary hypertension, and cardiac rehabilitation. She works as Referring MD for the High-intensity Cardiac Rehabilitation Unit at Multimedica in Milan.

Supplementary material

Supplementary material is available at European Heart Journal – Case Reports online.

Author contributions

Conceptualization, SDM and GC; validation, GC, SDM, AS, CB, and PA; writing—original draft preparation, SDM, GC, AS and CB; writing—review and editing, SDM, GC, CB, AS, and PA. All authors have read and agreed to the published version of the manuscript.

Silvia Di Marco (Conceptualization [lead], Data curation, Funding acquisition, Validation, Writing—original draft [equal]), Gaia Cattadori (Conceptualization, Data curation, Methodology, Validation, Writing—original draft, Writing—review & editing [equal]), Andrea Sonaglioni (Conceptualization, Data curation, Writing—original draft, Writing—review & editing [equal]), Carlo Banfi (Conceptualization, Data curation, Validation, Writing—original draft, Writing—review & editing [equal]), and Piergiuseppe Agostoni (Conceptualization, Data curation, Methodology, Supervision, Validation, Writing—original draft, Writing—review & editing [equal])

Consent: Informed consent was obtained from the patient in accordance with COPE guidelines.

Funding

This work has been supported by Ministry of Health - Ricerca Corrente to IRCCS MultiMedica.

Data availability

The data underlying this article are available partly in the article and its online Supplementary material, and partly upon reasonable request to the corresponding author.