CD4 + CD25 + CD127 high cells as a negative predictor of multiple organ failure in acute pancreatitis

Acute pancreatitis (AP) is an inflammatory disorder associated with high morbidity and mortality rate. The annual incidence of acute pancreatitis ranges from 13 to 45 per 100 000 people [1]. AP is the leading discharge diagnosis in patients admitted with gastrointestinal or liver problems in countries such as the United States [2]. Prognosis of the disease varies widely, from self-limiting to severe or critical, and mortality reaches 30% in severe cases [3].

Irrespective of the etiology of pancreatitis, early pathophysiology events include activation of digestive enzymes by lysosomal hydrolases, autodigestive processes, intra-parenchymal inflammatory response and the systemic inflammatory response syndrome (SIRS). The initial protease cascade doesn’t necessarily determine the severity of AP [4]. In contrast, evidence is accumulating of a crucial relationship between innate immune components involved in AP pathogenesis and disease severity [5‐7]. Neutrophils and macrophages are the first line of the immune system’s defense. T-lymphocytes, which are mainly involved in the cell-mediated immune response, also play a vital role in AP [5]. CD4 + T cells increase the severity of AP due to macrophage activation via antigen presentation and pro-inflammatory cytokine release as well as through direct cytotoxicity effects [5]. Immunological suppression mediated by regulatory T cells (Tregs) expressing transcription factor forkhead box P3 (FOXP3) has been reported to be the critical mechanism controlling the inflammatory response for which the immune system is primed after serious injury such as severe AP (SAP) [8]. However, these studies focused on a single lymphocyte subgroup rather than the broad subgroups of lymphocyte in patients with AP.

In present study, we evaluated the relationship between multiple peripheral lymphocyte subsets (i.e., T lymphocyte cells, T Helper cells, cytotoxic T cells, Tregs, activated effector T cells, natural killer (NK) cells and B cells) through flow cytometry done early during hospitalization vs. patient outcome. We then compared the accuracy of activated effector T cells estimation to predict non-MOF with Acute Physiology and Chronic Health Evaluation (APACHE) IIscores and C reactive protein (CRP) estimation of MOF.

This prospective clinical observational study was performed between June 2015 and August 2015 in the second affiliated hospital of Anhui Medicine University. The study was approved by the hospital’s Ethics Committee. Written informed consent was obtained from patient surrogates before study inclusion.

The principle findings of our study are: (1) among peripheral lymphocyte subsets, only activated effector T cells phenotyped by CD4 + CD25 + CD127^high have a significant negative correlation with MOF and can be used as a predictor of non-MOF estimation in AP; and (2) there is a statistically significant association between NK cells at admission and secondary infection.

AP can present as a severe acute abdominal disease with a high mortality. Patients with AP can develop an early-onset phase of SIRS within 2 weeks, and/or an infection phase due to pancreatic necrosis. MOF and/or infection signal the presence of severe disease, and there is a high risk of death if both are present [11]. Thus, both MOF and infection are unfavorable prognostic signs in AP patients.

In the present study, we did not find correlations between activated effector T cells and both SOFA scores and Balthazar scores (r = −0.276 p = 0.058 and r = −0.140 p = 0.342 respectively). In contrast, there were correlations between activated effector T cells and both APACHE IIscores and CRP. Furthermore, there were significant differences in the level of activated effector T cells among three subgroups stratified by Revision of Atlanta criteria (F = 5.26 p = 0.009), and the severe AP patients had the lowest level of activated effector T cells (5.36% ± 1.31%). From the result, it was demonstrated that activated effector T cells enrollment was in accordance with Ranking criteria, APACHE IIscores and CRP rather than Balthazar scores and SOFA scores. Balthazar scores based on the CT findings focus on local complications of AP and it is hard to reflect the systematic immunology condition. In terms of SOFA scores, it is not a normal clinical predictor in assessing severity of AP. considering the potential confounding bias caused by accumulative SOFA scores, it was just utilized for judging whether it was MOF individually in the present study.

An accurate predictor to prognosis would be helpful for clinicians by allowing patients with a predicted severe course to be transferred to an intensive care unit while patients with a predicted mild course may be treated in an outpatient setting. Severity scoring systems such as APACHEIIand laboratory testing such as CRP have been used to accurately predict an unfavorable outcome, which we have confirmed in the present study [12, 13]. However, there has not been a validated, published strategy to predict a favorable outcome, which is important to avoid overtreatment. In the present study, only a special T cell subset called CD4 + CD25 + CD127^high cells using the cutoff value >6.41% at admission was a statistically significant negative predictor of AP severity upon patient presentation to the hospital. Even though this predictor had a low sensitivity of 53.8%, we believe that the specificity of such a value is of more clinical importance than its sensitivity because conservative and supportive management is the initial care for patients with AP and it seems unnecessary to apply drastic remedies for those with higher CD4 + CD25 + CD127^high cells. This test’s potential as a superior predictor of a favorable prognosis will require confirmation in a larger population.

Lymphocytes play important roles in both adaptive and innate immune responses. The main function of the lymphocytic immune response is to mediate and resolve the nonspecific inflammatory process in AP. CD4+ T cells may act as co-stimulators for macrophage activation via antigen presentation and pro-inflammatory cytokine release as well as through direct cytotoxic effects (on acini) through Fas ligand expression on CD4+ T cells [14]. In addition, a distinct set of CD4 + CD25 + FOXP3 + Tregs presenting immunological tolerance and homeostasis have been identified with anti-inflammatory functions in AP [15]. These cells could bind to multiple effector immune cells and prevent their secretion of cytokines. They also secrete anti-inflammatory cytokines such as interleukin (IL)-10 and transforming growth factor (TGF)-β. Tregs have also been reported to restrain SIRS [16]. It was impossible to purify living Tregs in vivo previously because of the exclusive intracellular expression of FOXP3 as a critical factor for Tregs function. Recently, CD127 (IL-7 receptor a chain), whose expression was reverse to FOXP3, was recommended as an marker of Treg and has confirmed superiority compared to other cell surface markers [17, 18]. As indicated in previous study, serum sCD163, a biomarker released from macrophages, was increased in AP, but not associate with disease severity [19]. In present study, we used the sorting strategy of combining CD4+, CD25+, and CD127low to obtain viable, expandable Treg cells. There was no significant difference in Tregs between the two groups statistically, and we also found the mean number of Tregs as a percentage of CD4+ T cells was 5.55 ± 1.69%, which was a little lower than the 6.25 ± 0.26% in healthy people reported in previous research [18].

Prior studies have come to different conclusions regarding variation of Tregs in SIRS or the early phase of AP. Xue H et al. [20] reported that CD4 + CD25 + Tregs increased in SIRS and suppressed the excessive inflammatory response. In contrast, other investigators have reported that positive CD4 + CD25+ T cells significantly declined in a mouse SAP model [8]. Data on Tregs in humans with AP is limited and our data may be of value as a reference for future study.

Interestingly, the novel subgroup CD4 + CD25 + CD127^high cells had predictive value for the development of non-MOF in this investigation. This subset was previously thought to be a contaminant interfering with the suppressive function of Tregs and was neglected previously.[18] However, research by Michel L indicates that this specific population in multiple sclerosis (MS) appears more proliferative and secretes more interferon-γ (IFN-γ) and interleukin-2 (IL-2), both pro-inflammatory cytokines, than healthy individuals [21]. Similar observation has been made in CD4 + CD25 + CD127^high cells infiltrating rejected human allografts, in which allo-specific CD4 + CD25 + CD127^high cells were able to secrete effector cytokines such as tumor necrosis factor-α (TNF-α) and IFN-γ [22]. The ratio of CD127low/high also was statistically different in subgroups with and without MOF, but had no correlation with MOF in our multivariate regression model. Thus, CD4 + CD25 + CD127^high cells counts may be the sole factor which are associated with MOF estimation, and it is postulated that increment of CD4 + CD25 + CD127^high cells have a protective function to reflect the progress of AP independent of Tregs. The exact mechanism of CD4 + CD25 + CD127^high cell function in the setting of AP needs further study.

NK cells with phenotype marker CD3-CD16 + CD56+ are the type of cytotoxic lymphocytes critical to the innate immune system and secrete TNF-α and IFN-γ to control viral infection [23]. Dabrowski A et al. [24] found a dramatic depletion of peripheral NK cells in SAP compared to MAP and control, which indicates that there may be a correlation between NK cells and severity of AP. Our research further found that a lower number of NK cells at admission were associated with the development of secondary infection in AP patients. Since most pathogens causing secondary pancreatic infection are Gram-negative bacteria such as Escherichia coli originating from gut [25], we hypothesize that once circulating NK cells run into Gram-negative bacteria in a primary infection, they acquire memory function features which are a hallmark of T and B cells belonging to the adaptive immune system [26]. The memory-like, circulating NK cells then migrate into the pancreas at the early onset of AP and trigger a rapid immune response to the same antigen encountered again. In any case, patients with lower NK cells at admission may have a high risk of developing secondary infection in AP.

Increased peripheral CD4 + CD25 + CD127^high cells, whose physiological role in the early phase of AP is unclear in the early phase of AP, appear to be associated with a good (non-MOF) prognosis. In addition, patients with lower NK cells at admission may have a higher risk of developing secondary infection in AP. Further study is needed to confirm these observations.