Lymphopaenia in cardiac arrest patients

Despite the high mortality rates still associated with sudden cardiac arrest (CA), advances have been made in recent years, in particular to improve the proportion of patients achieving return of spontaneous circulation (ROSC) after cardiopulmonary resuscitation (CPR) [1]. Although the implementation of high-quality CPR and early defibrillation has increased the number of CA patients being admitted to the hospital, there has been no associated increase in survival rates and neurological recovery among CA survivors [2, 3]. The high mortality rate observed in this patient population is related to the consequences of the global ischaemia–reperfusion process, the so-called post-cardiac arrest syndrome (PCAS), which is characterised by a systemic inflammatory response that may be involved in the development of myocardial dysfunction, brain injury and multiple organ failure [3, 4].

The PCAS is a complex pathophysiological process, which encompasses a generalised activation of immunological and coagulation pathways and has many features common with sepsis [4]. Adrie et al. [4] reported a significant increase in the concentrations of various serum cytokines in CA patients in the first hours after the initial injury, which was more significant in non-survivors than in survivors. Other studies have also shown impaired microcirculatory perfusion and increased plasma endotoxin levels in CA patients, similar to findings in sepsis [5, 6]. Interestingly, sepsis is rapidly accompanied by a state of relative immunosuppression, characterised by a reduction in the number of functional immune cells and T-cell dysfunction [7, 8], which predicts a poor outcome [9].

Among the different components of the immune response, lymphocytes are essential in the organism’s defence against external aggression, by interaction (T cells) with antigen-presenting cells and subsequent activation of the immune response (B and T cells), in a complex interplay of cell-to-cell interactions, production of adhesion molecules and modulation of growth factors [10]. Severe lymphopaenia has been described in almost 30% of patients admitted to the intensive care unit (ICU) with severe sepsis or septic shock and was associated with high plasma levels of tumour necrosis factor-α, interleukin (IL)-6 and IL-10 [11]. Lymphopaenia during sepsis was also associated with features of immunosuppression, such as spontaneous hypothermia and an increased risk of hospital-acquired infections, and was an independent predictor of poor outcome [9, 12]. Persistent lymphopaenia has also been associated with poor outcomes in trauma patients, in particular when an infection occurred during the hospital stay [13]. Although survivors from CA are at high risk of secondary infections, in particular early-onset pneumonia [14], no data are available on the relationship between lymphocyte count and the risk of infection or patient outcome. Thus, the aim of our study was to evaluate the prevalence of lymphopaenia among CA survivors and its association with ICU survival and long-term neurological outcome. We hypothesised that lymphopaenia could be a significant predictor of poor outcome in this setting.

Of the 424 eligible patients admitted during the study period, 23 died within the first 24 h and 24 were excluded for missing data, so that 377 were included in the final analysis. The main characteristics of the study population (age 62 [52–72] years—OHCA 56%) are given in Table 1. Median time to ROSC was 15 min and 232 (62%) patients had a non-shockable initial rhythm; 150 (40%) patients had a non-cardiac origin of arrest. ICU mortality was 58% (n = 217) and 131 patients (38%) had a favourable neurological outcome.

In this study, 40% of post-CA patients had lymphopaenia, and in one-third of them, this persisted through the first 48 h after admission. Predictors of lymphopaenia on admission were older age, a shorter resuscitation time, a history of corticosteroid therapy and high levels of the inflammatory marker (e.g. CRP). Lymphopaenia was more frequent in patients with poor outcome but was not an independent predictor of mortality or unfavourable neurological outcome, unless in the subgroup of OHCA patients.

The occurrence of lymphopaenia and its impact on survivors of CA has not been well studied. In a pig model of prolonged CA and CPR, Gu et al. reported a high degree of splenic lymphocyte apoptosis, which was initiated by activation of the Bcl-2/Bax mitochondrial pathway [18]. These authors also described a reduction in CD4⁺/CD8⁺ T lymphocytes and a shift in these cells from an anti-inflammatory (Th-2) to a pro-inflammatory (T-helper [Th]-1) state in the myocardium [19, 20]. In mice submitted to global cerebral ischaemia secondary to CA within 3 h after resuscitation, Deng et al. [21] showed the presence of infiltrating peripheral CD4⁺ T lymphocytes in the brain. These experimental data indicate that early disturbances of immunological function after ROSC are associated with an increased production of inflammatory mediators and lymphocyte apoptosis, as well as an intense cardiac and neuroimmune response in which infiltrating T cells may play a key role. Few data are available in humans. In 50 OHCA patients, Venet et al. [22] reported a moderate decrease in the number of circulating CD4⁺ T lymphocytes although total lymphocyte count was normal. Nevertheless, the high mortality rate in that study (90%) prevented a comparison of lymphocyte levels in survivors and non-survivors. Indeed, as clinical studies have demonstrated that PCAS is characterised by high plasma cytokine and endotoxin levels, future studies should better characterise how these abnormalities may alter lymphocyte sub-populations and function in CA survivors and whether these changes are reversible or can be influenced by therapeutic interventions [23, 24].

Lymphopaenia has been widely described in patients with sepsis [12, 25, 26]. After an initial predominant pro-inflammatory phase, many septic patients develop persistent immunosuppression, characterised by increased inhibitory receptors on T cells and antigen-presenting cells, decreased production of pro-inflammatory cytokines, expansion of myeloid-derived suppressor cells, and apoptosis-related loss of T and B-lymphocytes and dendritic cells [7, 27]. Induction of lymphocyte apoptosis increased mortality and prevented bacteraemic control in septic animals [28]. The detrimental effects of apoptosis are not only related to the severe loss of immune cells but also the impact that apoptotic cell uptake has on the surviving immune cells. As such, uptake of apoptotic cells by monocytes, macrophages and dendritic cells results in immune tolerance and cellular anergy, which is associated with increased IL-10 production and the induction of a Th-2 cell immune phenotype; the net result of these changes is that the surviving phagocytic cells cannot provide adequate defence against infection [29].

In our study, patients with lymphopaenia were older than those without. Lymphopaenia is a common finding in elderly hospitalised patients and has been associated with poor outcome in these patients [30]. There is a significant interaction between the immune system and the ageing process, which may also influence the occurrence of chronic diseases; in particular, thymic demise represents an important phenomenon that can cause a reduction in T-cell count and peripheral proliferation of pre-existing T-cell clones, which can trigger limited immune reactivity damage-associated molecular patterns [31]. As expected, previous therapy with corticosteroids was also an independent determinant of lymphopaenia [32]. Data on the prevalence of lymphopaenia among corticosteroid users are scarce and biased by the presence of concomitant malignancies, administration of chemotherapy or other immunosuppressive drugs. The mechanisms include increased Th-2 cells with reduction in pro-inflammatory circulating cytokines or externalisation of phosphatidylserine, which may trigger cellular apoptosis, in particular of CD4⁺ cells [33]. Nevertheless, we found an association between increasing CRP levels and the risk of lymphopaenia. This is probably due to increased production of IL-6, which has been shown to be an independent predictor of poor outcome in CA patients [34]. In a recent study, the injection of endotoxin in healthy volunteers was associated with systemic inflammation, which triggered the occurrence of lymphopaenia, by an increase in anti-inflammatory regulatory T cells and a relative functional impairment of T-cell cytokine production, despite detectable levels of plasma pro-inflammatory cytokines [35]. Predictors of lymphopaenia in patients without previous immunosuppressive therapy were similar to those of the entire cohort, underlying again their relevance independently of patients’ immune status.

Interestingly, a shorter time to ROSC was another variable associated with lymphopaenia. We may have expected that the severity of the ischaemic injury, which is related to the longer duration of CPR, would have been associated with more severe apoptosis of T cells and have induced a low lymphocyte count in this setting. In stroke, although ischaemic cerebral damage may lead to suppression of peripheral immune responses, which predisposes to infection, no association was found between the extension of the infarct area and functional immune alterations [36]. The higher prevalence of IHCA in the lymphopaenia group with a shorter intervention time and CPR may also explain this finding. Additional studies should evaluate the complex interplay between pre-existing comorbid conditions and acute cerebral injury and the development of altered immune responses in CA patients.

There was no independent association between lymphopaenia and poor outcome. In patients with sepsis and septic shock, a low lymphocyte count was a strong predictor of poor outcome, even better than increased neutrophil cells [12]. Although sepsis has many common features with the PCAS, lymphocyte count and function may play a marginal prognostic role in CA survivors. One possible explanation could be that the severity of the initial anoxic brain injury (i.e. bystander CPR and time to ROSC) was not correlated with the occurrence of lymphopaenia. Moreover, many patients died within the first 2–3 days after arrest because of significant cardiovascular impairment or extended post-anoxic brain injury [37], limiting the impact of lymphopaenia on the occurrence of secondary infections or delayed organ dysfunction. Nevertheless, we observed that in the subgroup of OHCA patients, lymphopaenia on admission was associated with UO. This suggests that factors as the underlying different causes of the arrest between IHCA or OHCA or the management of CPR according to arrest location (i.e. different response time and quality of CPR) are an important determinant of biological biomarkers of poor prognosis and that routine monitor of the lymphocyte count in OHCA patients could be considered. However, future research should focus on a better characterisation of lymphocytes sub-populations, in relationship with other biomarker of “immunosuppression” (i.e. HLA-DR on monocytes) to better understand the potential prognostic and therapeutic role of lymphopaenia in this patients’ population.

Our study has some limitations. Firstly, it was a retrospective, single-centre study, which may limit the generalisability of our conclusions; this can, however, also be an advantage, as patients were treated according to a local protocol of PCAS management, thus reducing heterogeneity. Secondly, we did not record the exact time of no flow or the quality of CPR, which may influence the inflammatory response after reperfusion. Thirdly, all patients were treated with TTM so that we cannot draw any conclusions on lymphopaenia in normothermic CA patients or on the effects of TTM use, which may blunt the inflammatory response after rewarming, on the impact of lymphopaenia. Forth, causes of CA and lymphopaenia may differ between IHCA and OHCA. Also, it is possible that lymphopaenia may have preceded the arrest in some CA victims, such as in case of IHCA. Although location of arrest was not a significant predictor of lymphopaenia in our study, this might also be due to a lack of power and this question should be further addressed in future larger cohorts. Fifth, some patients, especially if suffering from IHCA, may have sepsis prior to ICU admission, and this could have been a significant trigger to lymphopaenia. Unfortunately, sepsis is widely under-recognised outside the ICU, so that this variable could not be reliably assessed in our database. Finally, we perform a multivariate model to identify variables associated with the occurrence of lymphopaenia, but we did not specifically investigate the lymphocyte count as a continuous variable. Although this might appear more appropriate, clinicians would be more interested in the presence of a “lymphopaenia”, its related risk factors and potential consequences rather than considering the absolute lymphocytes count.

Lymphopaenia is a common finding in CA survivors, in particular among those with an initial non-shockable rhythm, in-hospital CA and previous immunosuppressive therapy. A lower lymphocyte count on admission was associated with poor outcome, but was not an independent predictor of mortality or neurological recovery. Future studies should better characterise the immune response in patients resuscitated after CA to improve understanding of the pathophysiology of these findings and their potential therapeutic interventions.