Elevated Plasma Soluble PD-L1 Levels in Out-of-Hospital Cardiac Arrest Patients

Background: A deregulated immune system has been implicated in the pathogenesis of post-cardiac arrest syndrome (PCAS). A soluble form of programmed cell death-1 (PD-1) ligand (sPD-L1) has been found at increased levels in cancer and sustained inflammation, thereby deregulating immune functions. Here, we aim to study the possible involvement of sPD-L1 in PCAS. Methods: Thirty out-of-hospital cardiac arrest (OHCA) patients consecutively admitted to the ER of Mie University Hospital were prospectively enrolled. Plasma concentrations of sPD-L1 were measured by an enzyme-linked immunosorbent assay in blood samples of all 30 OHCA patients obtained during cardiopulmonary resuscitation (CPR). In 13 patients who achieved return-of-spontaneous-circulation (ROSC), sPD-L1 levels were also measured daily in the ICU. Results: The plasma concentrations of sPD-L1 in OHCA were significantly increased; in fact, to levels as high as those observed in sepsis. sPD-L1 levels during CPR correlated with reduced peripheral lymphocyte counts and increased C-reactive protein levels. Of 13 ROSC patients, 7 cases survived in the ICU for more than 4 days. A longitudinal analysis of sPD-L1 levels in the 7 ROSC cases revealed that sPD-L1 levels occurred in parallel with organ failure. Conclusions: This study suggests that ischemia- reperfusion during CPR may aberrantly activate immune and endothelial cells to release sPD-L1 into circulation, which may play a role in the pathogenesis of immune exhaustion and organ failures associated with PCAS.


Introduction
Post-cardiac arrest syndrome (PCAS), which occurs in resuscitated patients undergoing cardiac arrest, is characterized by four key pathological manifestations including post-cardiac arrest brain injury, post-cardiac arrest myocardial dysfunction, systemic ischemia-reperfusion, and persistent precipitating pathophysiology [1][2][3]. Successful management of PCAS constitutes a vital component of post-cardiac arrest care, critically affecting the prognosis of resuscitated out-of-hospital cardiac arrest (OHCA) patients [4]. An ischemia-reperfusion injury that systemically occurs in PCAS induces a deregulated immune response and sustained inflammation, resembling the pathophysiology of sepsis [5]. Immune-suppression represents an important pathology involved not only in sepsis, but also in PCAS, thereby increasing the risk of such infections as pneumonia during postcardiac arrest care [6]. Leukocytes isolated from PCAS patients have been shown to exhibit

Materials and Methods
This study protocol was reviewed and approved by the Institutional Review Board (IRB) of the Mie University Graduate School of Medicine (#3027). Informed consent to participate in this study was obtained in all cases from close family members of OHCA patients admitted to the emergency department and intensive care unit (ICU) at Mie University Hospital Japan.

Study Design and Patient Characteristics
This study enrolled 30 consecutive OHCA patients from August 2020 to May 2021. Upon arrival at the hospital, all patients were immediately subjected to our institute's standard cardiopulmonary resuscitation (CPR) protocol. Specifically, patients were intubated, manually ventilated with 100% oxygen, and treated with standard chest compressions by ICU physicians. The end-tidal carbon dioxide (EtCO2) concentrations were monitored to assure effective heart compression during CPR. Blood samples were drawn from the femoral arteries as soon as possible during CPR in all cases and from the arterial lines daily in the ICU in ROSC cases, and were used primarily for clinical purposes (e.g., laboratory testing). Parts of the blood samples were used to measure plasma cell-free PD-L1 concentrations. Platelet-free plasma fractions were prepared from blood samples as previously described [20]. Platelet free-plasma samples were kept at −80 • C until use. Patient clinical and laboratory data were extracted from electronic ICU records. In ROSC cases, the outcome from resuscitation was assessed by the cerebral performance category (CPC) score at the time of ICU exit. Whole-body, unenhanced postmortem computed tomography was performed for the cadavers of all patients to diagnose pneumonia.

Soluble PD-L1 Measurements
The concentrations of sPD-L1 in the plasma samples were measured using enzymelinked immunosorbent assay kits (Abcam, Japan, Cat#ab214565) according to the manufacturer's instructions with known concentrations of human recombinant sPD-L1 as an internal standard [19]. Each sample was analyzed in duplicate. The plasma levels of sPD-L1 in OHCA were compared with those of healthy volunteers and sepsis patients measured by exactly the same method in our previous report [19].

Statistical Analyses
Statistical analyses were performed using SPSS software v.25.0 (IBM Corp, Armonk, NY, USA). The results are presented as a median ± interquartile range, unless otherwise noted. Kruskal-Wallis tests were used to compare three or more groups. Mann-Whitney tests were used for two-group comparisons. To compare the correlations, Spearman's rank correlation was calculated between each data set. A p-value < 0.05 was considered statistically significant.

Plasma PD-L1 Levels in OHCA Patients Were as High as Those in Sepsis Patients
The thirty OHCA patients enrolled in this study included 12 females and 18 males who ranged from 48 to 92 years old (75.4 ± 14.2) ( Table S1). The causes of OHCA included pneumonia (8 cases), acute myocardial infarction (8 cases), thoracic aortic dissection (4 cases), chronic heart failure (4 cases), suicide by hanging (2 cases), myocarditis (1 case), subarachnoid hemorrhage (1 case), and hyperkalemia from chronic renal failure (1 case) and asphyxia (1 case) ( Table 1). Twenty-six were non-shockable OHCA. Although, return of spontaneous circulation (ROSC) occurred in 13 cases, of whom 7 survived in the ICU for more than 4 days. We have studied the plasma levels of sPD-L1 as early as possible during CPR before ROSC, in which chest compressions partially restored circulation, potentially triggering ischemia-reperfusion injury [21]. We have shown that the plasma levels of sPD-L1 in OHCA during CPR were increased ( Figure 1). Compared with the previous results that we reported using the exactly same methods [19], the PD-L1 concentrations in OHCA were significantly higher than those of healthy volunteers ( Figure 1) and were as high as those of the sepsis patients measured by the exactly same method in our previous report ( Figure 1). We have studied the plasma levels of sPD-L1 as early as possible during CPR before ROSC, in which chest compressions partially restored circulation, potentially triggering ischemia-reperfusion injury [21]. We have shown that the plasma levels of sPD-L1 in OHCA during CPR were increased ( Figure 1). Compared with the previous results that we reported using the exactly same methods [19], the PD-L1 concentrations in OHCA were significantly higher than those of healthy volunteers ( Figure 1) and were as high as those of the sepsis patients measured by the exactly same method in our previous report ( Figure 1). A sub-group analysis has shown that the presence of pneumonia was associated with increased sPD-L1 levels in OHCA during CPR (Figure 2A). Nevertheless, the sPD-L1 levels in the OHCA sub-group without pneumonia remained higher than those in healthy A sub-group analysis has shown that the presence of pneumonia was associated with increased sPD-L1 levels in OHCA during CPR (Figure 2A). Nevertheless, the sPD-L1 levels in the OHCA sub-group without pneumonia remained higher than those in healthy volunteers and SIRS without sepsis ( Figure S1). Application of bystander CPR might reduce the duration of whole-body ischemia, thereby possibly mitigating reperfusion injury. However, bystander CPR did not affect sPD-L1 levels in OHCA during CPR ( Figure 2B). The sPD-L1 levels in OHCA during CPR did not correlate with the subsequent occurrences of ROSC ( Figure 2C). volunteers and SIRS without sepsis ( Figure S1). Application of bystander CPR might reduce the duration of whole-body ischemia, thereby possibly mitigating reperfusion injury. However, bystander CPR did not affect sPD-L1 levels in OHCA during CPR ( Figure  2B). The sPD-L1 levels in OHCA during CPR did not correlate with the subsequent occurrences of ROSC ( Figure 2C).

Correlations of Plasma sPD-L1 Levels with Clinical Parameters in OHCA Patients
While investigating how sPD-L1 levels affected the clinical laboratory test results sampled during CPR, we found that they correlated with increased blood urea nitrogen (BUN) and creatinine (Cre) and decreased estimated glomerular filtration rate (eGFR) values (Table 2, Figure 3). In addition, sPD-L1 levels correlated with reduced lymphocyte numbers and increased C-reactive protein levels ( Table 2).

Correlations of Plasma sPD-L1 Levels with Clinical Parameters in OHCA Patients
While investigating how sPD-L1 levels affected the clinical laboratory test results sampled during CPR, we found that they correlated with increased blood urea nitrogen (BUN) and creatinine (Cre) and decreased estimated glomerular filtration rate (eGFR) values (Table 2, Figure 3). In addition, sPD-L1 levels correlated with reduced lymphocyte numbers and increased C-reactive protein levels ( Table 2).

Longitudinal Changes of sPD-L1 Levels in ROSC Cases
To examine the longitudinal changes of sPD-L1 levels in PCAS, we studied 7 ROSC patients who survived for more than 4 days in the ICU (Figure 4) and, thereby, sought to correlate the sPD-L1 levels with the severity of organ failures and the outcome. In these cases, we found that sPD-L1 levels remained higher than normal, and appeared to corre-

Longitudinal Changes of sPD-L1 Levels in ROSC Cases
To examine the longitudinal changes of sPD-L1 levels in PCAS, we studied 7 ROSC patients who survived for more than 4 days in the ICU (Figure 4) and, thereby, sought to correlate the sPD-L1 levels with the severity of organ failures and the outcome. In these cases, we found that sPD-L1 levels remained higher than normal, and appeared to correlate with the severity of organ failures as represented by the SOFA scores (Figure 4). The CPC score was used to assess the outcome of resuscitation. In the cases presenting high CPC scores [22] at the time of ICU discharge, sPD-L1 levels remained exceedingly high as did SOFA scores ( Figure 4A-C). By contrast, in other cases presenting low to moderate CPC scores at the time of ICU discharge, sPD-L1 levels gradually reduced, as did SOFA scores ( Figure 4E-G CPC score was used to assess the outcome of resuscitation. In the cases presenting high CPC scores [22] at the time of ICU discharge, sPD-L1 levels remained exceedingly high as did SOFA scores ( Figure 4A-C). By contrast, in other cases presenting low to moderate CPC scores at the time of ICU discharge, sPD-L1 levels gradually reduced, as did SOFA scores ( Figure 4E-G).

Discussion
In this pilot study involving 30 OHCA patients, we observed significantly elevated levels of plasma sPD-L1 during CPR. The sPD-L1 concentrations in OHCA were as high

Discussion
In this pilot study involving 30 OHCA patients, we observed significantly elevated levels of plasma sPD-L1 during CPR. The sPD-L1 concentrations in OHCA were as high as those of sepsis [19], thereby supporting our hypothesis that the PD-1/PD-L1 signaling pathway might be involved in the pathogenesis of both PCAS and sepsis [12]. The increased sPD-L1 levels are thought to result from ischemia-reperfusion, because, even before ROSC, circulatory support by CPR with chest compressions partially restored tissue perfusion, thereby activating the mechanisms that lead to ischemia-reperfusion injury [21]. Infections such as pneumonia [23] and underlying medical conditions present in the current OHCA cohort (Table 1) such as diabetes mellitus [17], rheumatoid arthritis [24], and systemic lupus erythematosus [25] have been shown to associate with a modest increase in the plasma sPD-L1 levels. Elevated sPD-L1 levels were found in all OHCA cases regardless of their underlying medical conditions, thereby supporting the impact of ischemia-reperfusion injury during CPR to cause a significant increase in sPD-L1 in the plasma. However, it is possible that pneumonia and some underlying medical conditions could contribute to the increased sPD-L1 levels seen in this study. Further investigations are needed to confirm these suspicions.
sPD-L1 levels during CPR were indistinguishably elevated in both ROSC (+) and ROSC (−) patients. To further study how sPD-L1 levels unfold during PCAS, we examined 7 ROSC patients treated in the ICU for more than 4 days. In all cases, sPD-L1 levels remained high, albeit changing in proportion to the SOFA scores. This preliminary finding needs to be further substantiated in future investigations to support our contention that sPD-L1 may be included as a biomarker candidate to predict the severity of organ failures in PCAS. Of note, in sepsis patients, sPD-L1 levels have been shown to correlate with SOFA scores, as well as with specific laboratory and clinical parameters of the impaired renal, coagulation, and central nervous systems [19].
PD-L1 is expressed not only in leukocytes such as macrophages and monocytes, but also in non-leukocytic cells such as endothelial and epithelial cells [15,26]. Although the present study did not demonstrate the upregulation of activation markers for endothelial cells and leukocytes, previous studies have shown that ischemia-reperfusion injury in PCAS induces aberrant endothelial [27][28][29] and leukocyte activation [30,31]. Thus, leukocytes and endothelial cells likely release sPD-L1 in resuscitated OHCA patients. sPD-L1 is mostly produced by proteolytic cleavage of the extracellular part of membrane-bound PD-L1, although a small proportion of sPD-L1 in the plasma may originate from alternatively spliced PD-L1 mRNA lacking a transmembrane domain [15]. Several proteases such as matrix metalloproteinase (MMP)-9 [32], MMP-13 [33], a disintegrin and metalloproteinase (ADAM)10 [34], and ADAM17 [35] have been shown to cleave membrane-bound PD-L1, thereby forming sPD-L1. Notably, MMPs and ADAMs, including MMP-9 [36], MMP-13 [37], ADAM10 [38], and ADAM17 [39], have been shown to be activated in ischemia-reperfusion injury models. Thus, one plausible scenario is that the ischemiareperfusion injury that occurs during CPR activates MMPs and ADAMs, which cleave off PD-L1 on leukocytes and endothelial cells to produce sPD-L1.
As sPD-L1 retains the ability to bind PD-1 on the cell surface, thereby transmitting signals to dampen immune cell activation [15], increased levels of sPD-L1 have been implicated in the perturbed anti-tumor immunity observed in cancer patients [11] and in the immune paralysis suffered by sepsis patients [19]. Increased sPD-L1 in OHCA patients may constitute an important component of the immunosuppressive milieu in the plasma, as previously suggested in cases of PCAS [5]. In this way, aberrant activation of PD-1 signaling by the increased expression of sPD-L1 could give rise to unwanted immune suppression, which could predispose patients to opportunistic infections [12]. Activation of PD-1 signaling by PD-L1 induces not only T cell unresponsiveness, but also apoptosis of CD4 [40] and CD8 [41] T cell subsets. PD-1 is also expressed in B cells [42] and activationinduced apoptosis of memory B cells has been reported in critically ill patients [43]. Thus, studying the potential depletion of different lymphocyte subsets in OHCA would be of great interest.
Alternatively, increased levels of sPD-L1 may have regulatory and protective properties, as the PD-1 signaling elicited by PD-L1 alleviates ischemia-reperfusion injuries to the kidney [44] and liver [45]. However, conflicting results have been reported in ischemiareperfusion in the brain. PD-1 deficiencies in knockout mice worsened brain infarction in cases of ischemia-reperfusion, supporting the protective role played by PD-1 signaling [46]. By contrast, PD-L1 deficiencies in knockout mice reduced brain infarction [47]. Thus, the roles of PD-1 signaling, as activated by PD-L1 during ischemia-reperfusion, may vary depending on the types of organs. This could partly be explained by the deregulated balance between effector and regulatory T cells in PCAS [48], as PD-1 signaling induces opposing effects in each [49]. Whereas PD-1 signaling in effector T-cell function suppresses inflammation, the signaling in regulatory T-cells results in the inhibition of immune-suppressive effects, thereby augmenting inflammation [49].
A potential limitation of the present study is the concern that age differences could explain the cause of the increased sPD-L1 levels in the OHCA cohort, as the OHCA cohort is older than the healthy volunteer cohort (Table S1). The previous study that measured plasma sPD-L1 concentrations in healthy volunteers of different age groups reported that the concentrations between the 31-50-year-old group and that of the 51-70-year-old group were similar [50], thereby suggesting a modest effect, if any, of ages in the present results. However, as age-related underlying medical conditions could also affect the sPD-L1 measurements, further investigations involving an increased number of OHCA patients and age-matched cases with similar underlying medical conditions are needed to address the concern. Another potential limitation is that one should be cautious in interpreting the correlations of sPD-L1 with certain clinical laboratory test results (e.g., increased BUN, increased Cre, and decreased lymphocyte counts) and increased C-reactive protein levels in blood samples obtained during CPR, as such test results may simply reflect the severity of pre-existing diseases and/or the severity of ischemia-reperfusion.

Conclusions
In summary, this study supports our hypothesis that elevated levels of sPD-L1 in OHCA patients during CPR reflect systemic ischemia-reperfusion, which aberrantly induces immune and endothelial cells to shed the extracellular part of PD-L1 into the circulation. Our results suggest that sPD-L1 may play a role in the pathogenesis of immune exhaustion and the multi-organ failures associated with PCAS. Thus, sPD-L1 could be added to the list of reported biomarkers of PCAS [21,28,30,[51][52][53][54][55]. Follow-up control studies involving a large cohort of OHCA patients are needed to validate our hypothesis.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/jcm10184188/s1, Figure S1. Comparisons of plasma sPD-L1 levels in the subgroup of OHCA without pneumonia, with sepsis, systemic inflammatory response syndrome (SIRS), and healthy volunteers; Table S1. Clinical parameters of study participants. Informed Consent Statement: Informed consent to participate in this study was obtained in all cases from close family members of OHCA patients admitted to the emergency department and intensive care unit (ICU) at Mie University Hospital Japan.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.