Sepsis-associated pathways segregate cancer groups

Sepsis is a potentially life-threatening complication caused by dysregulated host response to infection, often leading to organ failure and death. Estimated global burden of sepsis is more than 48.9 million people in 2017 with 11 million deaths [1]. Septic shock is the advanced stage of sepsis with metabolic dysregulation and uncontrolled hypotension. Several epidemiological studies have linked sepsis and cancer [2, 3]. Liu et al. [2] conducted an association study between sepsis and ensuing risk of cancer in elderly adult population of the United States, and observed that sepsis is significantly associated with increased risk for many cancers including chronic myeloid leukemia, myelodysplastic syndrome, acute myeloid leukemia, cancers of the cervix, liver, lung, rectum, colon. Another association study revealed 2.5 fold increased risk of sepsis in survivors of cancer in community-dwelling adults (the risk is increased up to 10 times in hospitalized cancer patients) [3]. Co-occurrence of cancer with sepsis is associated with higher mortality than sepsis alone without cancer [4]. On the other hand, sepsis is a common cause of death in critically ill patients with cancer, with high ICU and hospital mortality [5‐7]. Interestingly, mortality due to sepsis varies widely from 42 to 82% across cancer tissue types [8], suggesting varying likelihood of survival of patients suffering from different cancers.

This study is motivated by multiple shared features of septic shock (SS) and cancer. There is co-occurrence of the two entities, with synergistic effect on mortality. Both are associated with inflammation at some stage of the disease. Inflammation is well understood to promote malignant growth with participation of diverse immune cells and molecules, such as, cytokines [9]. Similarly, sepsis is understood as a non-resolving inflammatory response to infection that leads to organ dysfunction [10]. Both the diseases are associated with anaerobic metabolism with lactic acidosis being a hallmark of septic shock [9, 11]. In our earlier work, we have observed a cancer associated pathway to be significantly up-regulated in septic shock [12]. Additionally, there are previous reports on shared molecular changes in sepsis and cancer. Bergenfelz et al. [13] reported that Wnt5a induces immunosuppressive phenotype of macrophages in sepsis and breast cancer patients. HMGB1, a key late inflammatory mediator of systemic inflammatory response syndrome associated with bacterial sepsis, is also implicated in tumorigenesis and disease progression [14]. Muscle wasting - observed in patients with cancer, severe injury and sepsis - is associated with increased expression of several genes, particularly transcription factors and nuclear cofactors, regulating different proteolytic pathways [15]. Methodologically, all of these studies employed gene-level analysis, i.e., considering gene as the functional unit. On the other hand, a gene set or pathway represents coordinated molecular activity and represents a higher-order functional unit in a tissue or cell. Pathway-level analysis allows detection of a cumulative signal that is not accessible at the gene-level. To our knowledge, there is no report in literature on pathway-level comparison between sepsis and cancer. In the present study, we have performed unbiased analysis of SS and cancer datasets to discover shared patterns of pathway-level transcriptional alteration underlying the two illnesses.

The transcriptomic data for 23 studies (17 cancer and 6 SS) consisted of 687 patients with cancer and 445 with SS. The data were subjected to progressive analysis and pathway filtering (Fig. 1).

Comprehensive system-level analysis reveals shared transcriptomic response between septic shock (SS) and a subset of cancers (SLC). The SLC group predominantly consists of cancer of the upper GI tract, including head and neck, oesophagus, stomach, liver and biliary tract. The striking segregation of the SLC group with SS suggests shared elements of pathophysiology operational in these disorders. Sixty-six pathways differentially expressed in both SS and SLC represent critical biological processes, such as metabolism, immune response and protection against infection. Network analysis reveals the selected pathways as a core functional module (Fig. 3) of the genome-scale network dysregulated in both SS and SLC. Many of these processes, such as metabolism and inflammatory response are known to contribute to the pathogenesis of both sepsis and cancer [9].

The segregation of cancer into two groups (SLC and CA) prompted us to investigate the prospect of the selected pathways as a classifier. Machine learning-based validation of an independent data set proved that the pathways indeed consist of a robust sample-level signature of the cancer groups. It is thus possible to predict, with high accuracy, whether a patient belongs to a particular cancer group (SLC or not). With plummeting costs and increasing acceptability of clinical transcriptomics, this signature may be useful in future for patient stratification.

Similarity in the direction of change in pathway activity immediately evokes an explanation in terms of shared immunological response in either case. SLC are often associated with infection (i.e., human papilloma virus in “head and neck”, H. pylori in stomach, hepatitis viruses in liver). In fact, TCGA data analysed for the presence of viral sequence reads in RNA-seq datasets of several cancers have revealed infection status of the samples. Interestingly, the majority of the infected samples belong to the GI associated cancers including head and neck (HNSC), oesophagus (ESCA), stomach (STAD) and liver (LIHC) [25‐29]. Cao et al. (2016) [26] estimated a high percentage of infected samples in TCGA datasets for liver (21.2%), head and neck (16.6%), stomach (14.8%) and oesophagus (12%). It is known that sepsis starts out as an infection, but it is the uncontrolled host response that drives the phase of shock. Similarly, in cancers with an infectious origin, host response (e.g., inflammation) plays a role in malignant transformation [9]⁠. Pathways related to immune response are consistently differentially expressed in both SS and SLC. Although the time scale of pathogenesis is much shorter in SS compared to cancer, our finding calls for greater attention into the shared cellular processes underlying these two distinct clinical phenotypes.

The finding of pathway up-regulation associated with survival brings new insight into the relationship between the shared transcriptomic patterns and disease outcome. Sepsis is a highly lethal dysregulated host response to infection, and SLC, by extension, appears to share a part of that response. Since the pathways are selected on the basis of up-regulation in both SLC and SS (compared to control), a plausible explanation is that the pathway up-regulation is protective rather than detrimental to the human host. It is worth mentioning that viral integration or bacterial infection in a setting of cancer has been reported to favour survival in SLC cancers of oropharynx [30], liver [31] and kidney [32].

Alteration of intestinal permeability is known in both sepsis and cancer [33]. It may be speculated that anatomical proximity increases the likelihood of the liver and upper gastrointestinal tissue to be exposed to the translocated microbial products through a permeable gut, eliciting a host response in both SS and SLC. Significant association between survival and up-regulation of phagocytosis-associated pathways (i.e., Fc gamma R-mediated phagocytosis) lends support to this view. In general, sepsis is less lethal in patients with cancer of the digestive system than cancer of other organs [8]. The role of gut in survival from sepsis and cancer is an active area of research with translational potential.

Firstly, transcriptome-based systems biology approach segregates cancer into two groups (SLC and CA) based on similarity with SS. Secondly, the similarity is based on a set of pathways associated with pathogenesis of sepsis and cancer. These pathways form a robust signature of a novel cancer grouping. Lastly, up-regulation of the pathways is protective rather than detrimental to the patient. A mechanism of the shared protective host response is hypothesized in terms of immunocompetence induced by microbial products from the permeable gut. This work is the first step towards a systems biology-based patient stratification. It is hoped that future work in this direction shall generate actionable knowledge for clinical management of both cancer and septic shock.