Immune Deregulation in Sepsis and Septic Shock: Reversing Immune Paralysis by Targeting PD-1/PD-L1 Pathway

Abstract

Sepsis remains a major problem for human health worldwide, thereby manifesting high rates of morbidity and mortality. Sepsis, once understood as a monophasic sustained hyperinflammation, is currently recognized as a dysregulated host response to infection, with both hyperinflammation and immunoparalysis occurring simultaneously from the earliest stages of sepsis, involving multiple organ dysfunctions. Despite the recent progress in the understanding of the pathophysiology underlying sepsis, no specific treatment to restore immune dysregulation in sepsis has been validated in clinical trials. In recent years, treatment for immune checkpoints such as the programmed cell death protein 1/programmed death ligand (PD-1/PD-L) pathway in tumor-infiltrating T-lymphocytes has been successful in the field of cancer immune therapy. As immune-paralysis in sepsis involves exhausted T-lymphocytes, future clinical applications of checkpoint inhibitors for sepsis are expected. In addition, the functions of PD-1/PD-L on innate lymphoid cells and the role of exosomal forms of PD-L1 warrant further research. Looking back on the history of repeatedly failed clinical trials of immune modulatory therapies for sepsis, sepsis must be recognized as a difficult disease entity for performing clinical trials. A major obstacle that could prevent effective clinical trials of drug candidates is the disease complexity and heterogeneities; clinically diagnosed sepsis could contain multiple sepsis subgroups that suffer different levels of hyper-inflammation and immune-suppression in distinct organs. Thus, the selection of appropriate more homogenous sepsis subgroup is the key for testing the clinical efficacy of experimental therapies targeting specific pathways in either hyperinflammation and/or immunoparalysis. An emerging technology such as artificial intelligence (AI) may help to identify an immune paralysis subgroup who would best be treated by PD-1/PD-L1 pathway inhibitors.

Keywords: PD-1; PD-L; artificial intelligence; immune checkpoints inhibitors; immunomodulation; immunoparalysis; machine learning; sepsis - diagnostics.

Figure 1

Paradigm shift in sepsis. In 1991, sepsis was defined as Systemic Inflammatory Response Syndrome (SIRS), which was caused by microbial infections. This definition emphasizes the concept that systemic inflammation is the key to sepsis. During the initial phases of sepsis, inflammation originating in the innate immune system is enhanced by multiple pathways as “cytokine storm”. Then, a new theory has since emerged positing that immunosuppression following initial hyperinflammation, eventually leading to prolonged and significant immunosuppression is the key pathophysiology. In terms of paired words, SIRS and compensatory anti-inflammatory response syndrome (CARS) symbolize this paradigm. In 2016, The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3) defined sepsis as life-threatening organ dysfunction caused by a dysregulated host response to infection. Both pro-inflammatory and anti-inflammatory genomic storms occur beginning in the earliest stages of infection, and the balance between the two determines whether clinically over-inflammation or immunosuppression then occurs. Persistent Inflammation, Immunosuppression, and Catabolism Syndrome (PICS) is an intriguing concept from the integrated point of view, which contends that SIRS and its counterpart CARS do not exist independently; rather, both are occurring simultaneously. As shown in the Surviving Sepsis Campaign guidelines, initial fluid resuscitation, earlier antibiotic administration and supportive care such as mechanical ventilation strategies, nutrition and PADIS (Pain, Agitation, Delirium, Immobility, and Sleep) management are key of sepsis management. With the development of these conventional basic treatment, there is an overall clinical outcome improvement, but mortality of sepsis still reaches high. Therefore, additional treatment method that targets the underlying essence of sepsis has been expected. At first, as sepsis was essentially understood as hyperinflammation, many anti-inflammation approaches were tried. However, none have demonstrated that those sepsis treatment strategies are effective. Then, focusing on the immunoparalytic aspect of sepsis, immune stimulation represents a new strategy for targeting sepsis. There are some promising molecules, among them PD-1/PD-L inhibitors, which can not only reverse immuostimulation but act as immunomodulation, are highly expected.

Figure 2

Subgrouping difference between human and Artificial Intelligence (AI). The reason why clinical trials have not been successful regardless of a deeper understanding of the pathology underlying sepsis is the diverse pathophysiology of sepsis in humans compared to animal sepsis models which are developed using relatively unified methods. Physicians usually classify septic patients by based on the types of primary disease, organ damage, and severity of systematic conditions that are consistent with clinical impressions. However, it is extremely difficult for even a skilled clinician to classify sepsis based on the limited information typically available at the beginning of treatment: medical history, vital signs, and few blood examinations. Thus, clinical trials could not have targeted specific patients. On the other hands, AI may achieve new classification, which is not transparent at first glance and is difficult for humans to understand, by machine learning. Only patients with immunoparalysis detected by AI should be treated with immune checkpoints inhibitors for successful clinical trials, for example.