By: Yichen Zeng
One of the fundamental problems of the human immune system lies in the question: “How many lives can one claim before one begins losing their own?” That is, how aggressive can the human immune system be towards pathogens and other invasive microorganisms before it begins to destroy the very being that depends entirely on it for survival?
Last year, the side effects of a revolutionary and versatile cancer treatment named “CPI therapy” prompted scientists to revisit this question when researchers at UCSF observed in CPI-treated patients the emergence of autoimmune diabetes. Though this complication is rare, occuring in only 0.2% to 1.9% of all CPI-treated patients, researchers believe the disease to be potentially life-threatening and as such, worthy of extensive further research – especially in light of how incredibly effective of a treatment CPI-therapy otherwise is.
CPI therapy, in essence, dials the aggression of the human immune system to eleven: through the use of chemical compounds called “checkpoint inhibitors” (CPIs) to prevent checkpoint molecules in the human body from impeding the activation of the immune system’s T cells, CPI therapy significantly accelerates the rate at which tumor cells are destroyed. However, while berserk immune cells may be welcome from a cancer treatment standpoint, they may in their frenzy inadvertently attack friendly cells – in CPI therapy’s case, the pancreatic beta cells crucial for maintaining blood sugar levels. Disruption to these pancreatic cells results in a disruption to blood sugar levels – in other words, diabetes or, specifically, CPI-induced diabetes. Yet, the mechanisms behind this CPI-induced disruption remain unclear.
One of the checkpoint molecules that CPI therapy inhibits – PD-L1 – has been proven to prevent diabetes in mice, and scientists suspect that there is a relationship between the expression of PD-L1 and the destruction of pancreatic beta cells following CPI therapy. To investigate this relationship, researchers at UCSF began by noticing that the blood lipase levels of CPI-treated patients who developed CPI-induced diabetes after treatment were significantly higher than those of patients who did not develop the disease. Since blood lipase levels correlate with the degree of inflammation of the pancreas, the three-fold increase in blood lipase levels that the researchers discovered in the patients who developed diabetes confirmed that pancreatic inflammation has a role in the development of this disease. Why inflammation arises alongside CPI-induced diabetes remains unknown.
Next, the research team proceeded to study the pancreatic tissue of a patient who had passed away from CPI-induced diabetes. In this patient’s pancreatic beta cells, they encountered a significant presence of PD-LI – to be expected, as PD-L1 expression tends to increase as diabetes progresses – and, after comparing these cells to ones from other studies, were able to confirm that this increased expression was related to the pancreatic inflammation as opposed to some other unknown side effect of CPI therapy. What’s particularly notable about this finding is that PD-L1 is often secreted in response to inflammatory mediators, specifically the cytokine IFN-γ known to be involved with pancreatic inflammation. With this lead, the research team decided to stain these pancreatic cells in order to confirm the presence of these cytokines, and they did – locating them within immune cells (“immune infiltrates”) that had infiltrated into the region of the pancreas – the “islets” – that houses the beta cells.
Is there more behind CPI-DM than merely the now-uninhibited infiltration of immune cells into the islets? First, the researchers had to determine whether or not autoimmune complications are characteristic of all pancreatic inflammations or only those due to CPI therapy. One series of experiments that they ran with test mice involved one mice group treated with anti-PD-L1 treatments and another group treated with CPI therapy that inhibited other checkpoint molecules. It was observed that mice from both groups developed immune infiltrates, but only mice treated with anti-PD-L1 treatments developed diabetes, indicating that anti-PD-L1 therapy was doing more than merely compelling immune cells to wreak havoc in the pancreas.
By analyzing the gene expression of immune and islet cells in the mice that were treated with anti-PD-L1 therapy through the use of qPCR, the researchers discovered that the IFN-γ response pathways in macrophages and other immune cells became enriched after anti-PD-L1 therapy, resulting in the immune cells’ becoming cytotoxic to pancreatic beta cells. What’s worse, however, is that in the pancreatic beta cells themselves, IFN-γ also increases gene expression of the production of other cytokines while decreasing the secretion of insulin – vital for maintaining blood sugar levels. Even worse, IFN-γ increases gene expression of pathways for cell death in beta cells.
At last, the comprehensive picture of the phenomenon at hand was relatively clear. Anti-PD-L1 treatments in CPI-therapy causes pancreatic inflammation, implying an increase in concentration of cytokines including IFN-γ which, in turn, goads beta cells into cell death among other adverse behavior and causes immune cells to infiltrate and damage the pancreas’s islets. While this is a rather bleak picture that casts doubt on the future of CPI therapy, not all hope is lost. Through administering anti-IFN-γ treatments among other anti-cytokine treatments to mice treated beforehand with anti-PD-L1 CPI therapy, these researchers observed a pronounced delay in the development of CPI-induced diabetes, bolstering the hypothesis that inhibiting cytokine levels after CPI therapy impacts the degree to which immune cells infiltrate the pancreatic islets.
However, much about CPI-induced diabetes remains unknown: for instance, it remains unclear why anti-PD-L1 treatments in particular are so prone to begetting complications in comparison to other checkpoint inhibitor treatments. Nonetheless, UCSF’s research on CPI-induced diabetes has shed much light on how this disease emerges, culminating in a study that yielded insight into strategies for designing potential treatments for this disease. And, with how promising of a cancer treatment CPI therapy has otherwise proven itself to be, it is understandable that we ought to devote as much of our resources as possible towards refining and perfecting this treatment method so that it may, one day, lead both the cancer research community and the greater medical field into another breakthrough in our understanding of this elusive malady.
