Acute cardiogenic or non-cardiogenic pulmonary edema sometimes develops in patients undergoing major cardiovascular surgery, although most cases present with bilateral pulmonary edema (BPE) [5]. Reportedly, UPE, which accounts for 2.1% of cardiogenic pulmonary edema (CPE) cases, usually occurs in the right lung [6]. Although there is no specifically established treatment for UPE that differs from BPE, UPE is associated with more severe conditions requiring mechanical ventilation and more intensive inotropic support [6]. The major cause of cardiogenic UPE is MR due to its propensity to flow anatomically into the right superior pulmonary vein [7]. Another underlying mechanism of right-sided UPE is a difference in interstitial flow in the lungs due to the right-sided lymphatic vessels having a smaller aperture and a longer distance to the thoracic duct than left-sided lymphatic vessels [8, 9].

In our patient, UPE occurred in the right lung. However, CPE was unlikely, since TEE did not reveal significant MR or LA enlargement, and instead revealed well-preserved left ventricular wall motion, while the PAC revealed normal PAP and CI, and chest X-ray revealed neither the increased cardiothoracic ratio nor “butterfly pattern” typically seen in CPE.

UPE can develop also due to other pathological conditions, such as acute pulmonary emboli, re-expansion pulmonary edema (REPE), and external compression of the pulmonary vasculatures by a tumor or hematoma, resulting in asymmetrical inflow to the pulmonary arteries or outflow to the pulmonary veins [10]. In our case, the right lung was left collapsed during CPB after discontinuation of mechanical ventilation, since the right chest cavity was opened during RITA graft harvesting. Furthermore, blood retention in the cavity might have aggravated the collapse. Therefore, REPE due to rapid re-expansion of the right lung following its collapse likely caused right-sided UPE. Typically, REPE develops only after the lung collapsed for 72 h or longer is rapidly reinflated [11]. However, since on-pump MICS via right mini-thoracotomies using one-lung ventilation are increasingly being performed, it has become evident that right-sided unilateral REPE can develop much earlier after CPB [12, 13]. Lung collapse is associated with the sequestration of leucocytes, and inflammatory activity is triggered when oxygen is supplied during re-expansion and concomitant reperfusion of the lung [14, 15]. This inflammatory response might be enhanced by multiple factors during cardiac surgery [12]. The inflammatory response to re-expansion and reperfusion can be aggravated by the generalized inflammatory response generated by CPB [12], as suggested by the association between prolonged CPB time and the development of UPE in MICS [13, 14]. The FiO2 set at 1.0 upon resuming ventilation might have worsened the inflammatory reaction during re-expansion, though we reduced the FiO2 to 0.5 immediately after measuring blood gas 5 min after CPB. Furthermore, the restricted bronchial artery blood flow during CPB might worsen lung ischemia [16]. Therefore, the combination of right lung collapse and prolonged CPB time might have resulted in the unilateral REPE in our case, although unilateral REPE has rarely been reported in patients undergoing conventional on-pump cardiac surgery [4], unlike those undergoing on-pump MICS [12, 13].

Since UPE occurred after the start of blood transfusion in our patient, we also considered the possibility of transfusion-related acute lung injury (TRALI) [17,18,19]. However, anti-HLA antibody or anti-granulocyte antibody, which reportedly triggers lung microvascular endothelial damage [19], was not detected in samples of transfused blood products. Additionally, TRALI usually presents with BPE [18]. Therefore, TRALI less likely caused UPE than REPE in our case.

The use of ECMO is indicated when a poor therapeutic response to conventional mechanical ventilation and a high mortality rate are predicted in patients with reversible lung injury [20, 21]. When considering the indication for ECMO, it is recommended to estimate the Murray score, in which a score of 2.5 or more indicates “severe” lung injury [21, 22]. Our patient was indicated for ECMO, since he was highly likely to die from severe hypoxemia without using ECMO, given his Murray score estimated to be 2.75 and PaO2/FiO2 as low as 42 mmHg.

In conclusion, we reported a rare case of UPE with severe hypoxemia that developed after conventional on-pump cardiac surgery. REPE seemed the likely cause of the UPE. V-V ECMO was quite effective in improving oxygenation and saving the patient’s life.