When enrolling patients who have undergone PCCS and are currently receiving HT, several important points need to be considered. We routinely performed pericardial closure during the initial surgery or used anti-adhesive patches when pericardial tissue was insufficient. Preoperative CT for re-sternotomy risk assessment was always performed. Although it may not be possible to cease anticoagulation and antiplatelet medications before HT owing to the unpredictable timing of organ availability [17], we promptly evaluated the candidate’s medication profile and initially suspended any potentially harmful drugs. As a preparatory measure before re-sternotomy, we routinely exposed the femoral artery and vein as an emergency route for CPB setup. To avoid the need for emergent CPB, it is crucial to allow an adequate amount of time for the surgeon to perform dissection. Continuous communication between the donor organ harvest team and the recipient preparation team is necessary to minimize CPB and allograft ischemia time. After confirming the suitability of the donor heart, the recipient team performed re-sternotomy. In urgent situations such as unexpected bleeding or changes in the donor’s condition, rapid CPB is established via femoral access for quick heart decompression and re-entry. After surgery, sternal closure may be delayed for 24 h if adequate hemostasis is not achieved. Furthermore, immunosuppressant and desensitization protocols were followed as usual, based on previous publications [18–21].

All statistical analyses were performed using the R software (version 4.1.0; R Foundation for Statistical Computing, Vienna, Austria). For descriptive statistics, means and standard deviations were calculated for nonparametric data. For categorical variables, numbers and frequencies were described. Mann–Whitney U, Fisher’s exact, and chi-square tests were used to compare the clinical characteristics and outcomes between patients with and without PCCS. Cox proportional analysis was used to identify independent factors associated with 1 year mortality and included all significant predictors in the multivariate analysis, with a p-value <0.05. Survival curves were plotted using Kaplan–Meier analysis; survival rates between patients with and without PCCS were compared using the log-rank test, and p-values <0.05 were considered statistically significant.

A total of 368 patients were included in this study, of whom 105 (29%) had PCCS and 263 (71%) did not. In the PCCS group, 53 patients (50%) underwent surgery for ICM and 52 (50%) for non-ICM. The range of timing between prior cardiac surgery and HT in patients with PCCS varied from 1 week to 44 years, with a median of 45 weeks (IQR 10–132 weeks). Detailed information on the PCCS type is shown in Figure 1.

The demographic data of the patients are shown in Table 1. Most heart failure cases in the PCCS group were due to ICM (50%), whereas dilated cardiomyopathy accounted for 65% of the cases in the non-PCCS group. Patients with PCCS were older, had a lower body weight, and had a higher incidence of cardiopulmonary resuscitation history (17% vs. 9%, p = 0.044), pre-transplant ventilator use, and extracorporeal membrane oxygenation (ECMO) support (19% vs. 11%, p = 0.039). During surgery, patients with PCCS had significantly longer CPB times than those without PCCS (220 ± 80 vs. 136 ± 49 min, p < 0.001). There were no significant differences between the groups regarding sex, blood type, UNOS status, pre-transplant intra-aortic balloon pump use, pre-transplant dialysis, diabetes, renal and liver function, age and body weight of donors, or allograft ischemic time.

Patients with PCCS had a higher incidence of post-transplant ECMO support (29% vs. 15%, p = 0.003), renal dialysis (41% vs. 21%; p < 0.001), and postoperative re-exploration or delayed sternal closure (31% vs. 18%; p = 0.004) (Table 2). Although not statistically significant, patients with PCCS also showed a higher incidence of early bloodstream infection and a longer post-transplant hospital stay (18% vs. 11%, p = 0.059; 49 ± 55 vs. 38 ± 23 days, p = 0.085, respectively).

The in-hospital mortality rate was significantly higher in patients with PCCS than in those without (26% vs. 7%, p < 0.001), and the 1 year mortality rate was also higher in the PCCS group (30% vs. 14%, p < 0.001). The leading cause of 1 year mortality in both groups was infection, with cerebrovascular events accounting for a higher proportion in the PCCS group (15.6% vs. 2.6%).

Table 3 presents the results of the Cox regression analysis conducted to identify risk factors for 1 year mortality. Univariate analysis revealed several variables associated with 1 year mortality, including recipient age, previous cardiopulmonary resuscitation, UNOS status, pre-transplant ventilator use, pre-transplant intra-aortic balloon pump support, pre-transplant ECMO support, creatinine level, pre-transplant renal replacement therapy, PCCS, donor age, and CPB time. However, in multivariate analysis, only PCCS (hazard ratio (HR) = 2.485, 95% confidence interval (CI) = 1.241–4.975, p = 0.01) and CPB time (HR = 1.005, 95% CI = 1.000–1.009, p = 0.044) emerged as significant risk factors for 1 year mortality.

In examining the influence of etiology on heart failure and the various types of surgery, univariate analysis revealed that ICM posed a significant risk factor for 1 year mortality compared to dilated cardiomyopathy (HR = 1.737, 95% CI = 1.037–2.911, p = 0.036). Additionally, coronary artery bypass grafting (CABG) was a substantial risk factor for 1 year mortality when compared to patients without prior cardiac surgery (HR = 3.391, 95% CI = 1.934–5.946, p < 0.001). As there was a strong correlation between the etiology of heart failure and the type of prior cardiac surgery, a multivariate analysis was conducted, categorizing patients into four groups based on the presence or absence of prior cardiac surgery and the etiology of ICM or non-ICM. After adjusting for other significant factors, the multivariate analysis demonstrated that patients with prior cardiac surgery for ICM had a 4.848-fold increased risk (95% CI = 1.644–14.299, p = 0.004), while patients with prior cardiac surgery for non-ICM had a 3.554-fold increased risk (95% CI = 1.016–12.439, p = 0.047) compared to those without prior cardiac surgery and non-ICM as the etiology.

All patients had complete follow-up data, with a mean follow-up duration of 7.1 ± 5.6 years. Kaplan–Meier survival analysis revealed lower 1, 5, and 10 years survival rates in the PCCS group than in the non-PCCS group (69.5% ± 4.5% vs. 85.6% ± 2.1%, 49.5% ± 5.0% vs. 71.7% ± 2.8%, and 39.2% ± 5.1% vs. 51.8% ± 3.3%, respectively; log-rank test, p = 0.0024, Figure 2A). However, excluding patients who died within the first year, the conditional Kaplan–Meier survival curve did not show a significant difference between the PCCS and non-PCCS groups (log-rank test, p = 0.33, Figure 2B).

The results of subgroup analyses for long-term survival are shown in Figure 3.

The UNOS database report shows that 30% of HTs were performed in recipients with prior cardiac surgery and that prior surgery increased the 1 year and 5 years mortality rates by 1.192 times and 1.104 times, respectively [12, 13, 22]. However, not all types of prior cardiac surgery have the same impact on HT outcomes. Studies indicate that VAD implantation does not affect the subsequent HT outcome, whereas re-transplantation for prior allograft failure exhibits a lower survival rate than other surgery types [11–14]. This study excluded patients with prior VAD implantation or heart re-transplantation to specifically examine the influence of PCCS on HT outcomes. Remarkably, patients with PCCS had a significantly higher 1 year mortality rate (30% vs. 14%) than those without PCCS. PCCS increased mortality risk by 2.485 times within the first year. Notably, the findings of this study highlight the amplified effect of prior cardiac surgery on HT outcomes when exclusively focusing on conventional cardiac surgery, emphasizing the necessity of meticulously considering this factor when selecting HT candidates.

Infection remained the leading cause of mortality in both groups in this study. Our previous research found that post-transplant dialysis and early bloodstream infection contributes to a 5.5-fold and 3.43-fold increase in early mortality, respectively [18, 21]. In the present study, the PCCS group exhibited a higher incidence of certain factors, namely, post-transplant ECMO support (29%), delayed sternum closure (31%), early bloodstream infection (18%), and dialysis (41%).

While previous studies have shown a correlation between prior cardiac surgery and an increased incidence of acute rejection after HT [6], immune sensitization presents a challenge for patients who have undergone prolonged VAD support or prior HT, which can negatively affect transplant outcomes [15, 16, 23]. However, after excluding these two high-risk groups, our study findings suggest that PCCS may not significantly elevate the risk of acute rejection following HT (7.4% vs. 16.7% in the groups with and without PCCS, respectively).

Interestingly, a significantly higher incidence of cerebrovascular event-related death, including hemorrhagic and ischemic stroke, was observed in the PCCS group (18.5% vs. 0% in the without PCCS group). This finding is consistent with a recent UNOS report that showed a significant increase in post-HT stroke in patients with prior cardiac surgery [24]. Of the five post-transplant strokes observed in this study, two were hemorrhagic, and three were ischemic. Three ischemic strokes occurred in recipients with ICM and prior CABG, who also received ECMO support before HT. ICMP, ECMO support, and re-sternotomy are recognized as the risk factors for post-HT stroke [24–26]. In this study, 50% of patients with PCCS received HT for ICMP, and 19% of patients with PCCS required preoperative ECMO support, which could explain the high incidence of stroke in this group. These findings suggest the importance of further cerebrovascular evaluation before HT, especially in patients with PCCS who require ECMO support and have an underlying ICMP.

In this study, the 10 years survival rate of patients with PCCS was 39%, significantly lower than that of the patients without PCCS. However, after excluding those who died within a year, the conditional survival analysis showed no significant difference in long-term survival outcomes between the groups. This indicates that the elevated mortality risk associated with PCCS primarily affects the early post-transplant period, consistent with previous studies [3–6, 27].

To evaluate the etiology and related surgery on HT outcome, the subgroup analysis revealed that patients with ICM and PCCS had significantly worse outcomes than those in other subgroups during the early postoperative period (Figure 3C). Interestingly, patients with ICM as the etiology but without PCCS showed good short-term survival; however, their long-term outcomes were as poor as those with PCCS. This finding aligns with UNOS reports indicating that ICM, as the etiology of heart failure itself, is a significant risk factor for poor survival after HT [22]. In the context of end-stage heart failure due to ICM, it remains debatable whether patients would benefit more from high-risk conventional bypass surgery for complete revascularization, or from medical treatment with VAD bridging to HT. Further research is needed to resolve this issue. Furthermore, although not statistically significant, patients with PCCS of non-ICM etiology showed worse survival outcomes in the mid-term follow-up (between 3 and 5 years after HT). It is important to consider the impact of different etiologies of heart failure, such as restrictive cardiomyopathy and rheumatic heart disease, on post-transplant outcomes [28–30]. Although our study attempted to address this impact, the limited number of cases prevented us from conducting a comprehensive analysis. Future studies utilizing large databases, such as UNOS reports, could provide more insights into this matter.

Our study also found that both PCCS and extended CPB time were significant risk factors of 1 year mortality. Despite no significant difference in cold ischemic time, patients with PCCS had a total CPB time that was an hour longer. Prolonged CPB can increase micro-emboli formation and the occurrence of renal and neurological complications [31, 32]. The longer duration between PCCS and re-sternotomy may reduce the difficulty of performing re-sternotomy [33]. We found a negative correlation between CPB time and the timing of PCCS and HT (R = −0.19, p = 0.05). However, the timing of PCCS to HT did not demonstrate a significant risk reduction in HT outcomes according to the univariate Cox survival analysis. Subgroup analysis also demonstrated that PCCS within or after 2 years did not significantly impact survival outcomes based on Kaplan–Meier analysis (Figure 3B). We found that the era in which HT took place significantly affected survival rates, with patients undergoing HT after 2010 exhibiting better survival outcomes (Figure 3A), thus reflecting advancements in perioperative care and growing HT experience. However, there were no significant differences in survival outcomes between patients with PCCS across different eras. It is important to note that re-sternotomy and prolonged CPB continue to pose challenges for HT in patients with PCCS. Allowing the surgeon sufficient time to perform a demanding re-sternotomy without needing CPB is crucial for improving HT outcomes.

The limitations of this study include the small sample size and the nearly three-decade span it covers, during which significant changes in cardiogenic shock, AKI, and HT management have occurred. While we conducted a thorough examination within our center, it is important to recognize that the findings may have a more localized impact, potentially being more applicable to a single-center scenario rather than offering a comprehensive analysis suitable for a wider range of centers. To better understand the effects of various subgroups, such as the timing of PCCS, type of surgery, and effect of mechanical support (e.g., durable and non-durable VAD support or re-transplantation), a larger number of patients would need to be recruited within a shorter timeframe. This could be achieved through multicenter collaboration or by analyzing a national registry.

Despite these limitations, our study underscores the increased risks and complications faced by patients with PCCS undergoing HT. Therefore, meticulous patient selection and management strategies, including preoperative assessment of re-sternotomy risks, are vital for improving the outcomes. Further research is required to investigate whether the use of VAD in patients with PCCS would improve outcomes after HT and to determine any potential benefits it may offer.