As the world’s elderly population has grown, the number of patients with chronic kidney disease (CKD) has increased in recent decades. Multiple factors, such as kidney aging, glomerulonephritis and comorbidities, influence the progression of CKD in the elderly. While many innovative therapies for CKD have been developed, the mechanism of kidney aging remains unclear and its understanding has recently become an emerging topic.

In kidneys, aging leads to progressive nephron loss and increased glomerulosclerosis. Podocyte senescence plays a key role in this age-related glomerulosclerosis: Recent studies have shown that podocyte loss is associated with kidney aging (1) and that overexpression of programed cell death-1 (PD-1) in aged podocytes leads to cell death by interacting with its ligand, programed cell death ligand 1 (PD-L1), resulting in podocyte loss (2). As immune checkpoint molecules, PD-1 and PD-L1 play crucial roles in maintaining immune homeostasis. A representative example of this interaction is the blockade of anti-tumor immunity; PD-L1 on the membrane of cancer cells suppresses anti-tumor immunity by binding to the PD-1 receptor primarily on activated T cells (3). Previous studies have shown that in inflammatory kidney diseases where T cells play a prominent role in immune responses, such as allograft rejections, glomerulonephritis and tubulointerstitial nephritis, PD-L1 is overexpressed on renal tubular and vascular epithelial cells, suggesting the possibility of its renoprotective function (4–6). In kidney aging, the interaction between PD-1 and PD-L1 has recently drown great attention. Pippin et al has demonstrated overexpression of PD-1 in aged murine and human podocytes, which correlated with lower eGFR, higher segmental glomerulosclerosis, and vascular damage (2, 7). In turn, blocking PD-1 with an antibody inhibited apoptosis of podocytes and improved phenotype aging in murine kidneys (2). This activation of PD-1 signaling needs its ligand PD-L1, but it has not been determined how PD-L1 initiates the signaling in podocytes. Similar to its membrane-bound form, soluble PD-L1 (sPD-L1) in the blood is thought to exert inhibitory effects on immune function by binding to PD-1 on T cells (8, 9), and elevated blood levels of sPD-L1 are reportedly associated with various conditions, such as cancer (10, 11), aging (12), pregnancy (13), and diseases (8), e.g., acute coronary syndrome (14), severe viral infections, e.g., COVID-19 (15), and severe sepsis (16).

It is possible that the circulating sPD-L1 also affects the PD-1 signaling in podocytes. Thus, aging is hypothesized to elevate sPD-L1 levels in blood, as well as PD-1 expression on podocytes, potentially facilitating podocyte death via interaction between these molecules, thereby advancing CKD stages.

It is important to have an understanding of the relationship between serum sPD-L1 levels and kidney function before proceeding with detailed investigations, which is the aim of this study.

In this study, serum sPD-L1 levels were found to increase exponentially with CKD stage progression, independent of age, sex, or other confounding factors, a finding further validated by assessing another cohort. Although numerous studies have evaluated sPD-L1, few have described its association with kidney function. sPD-L1 is produced by either alternative splicing or proteolytic cleavage of membrane PD-L1, and has a molecular weight of approximately 37–56 kDa (8). sPD-L1 might be excreted in urine (20, 21). Thus, serum sPD-L1 levels increase with reduction of GFR, as evidenced by the fact that sPD-L1 levels increased with CKD progression in the present study.

Furthermore, median sPD-L1 levels were found to be three times higher in dialysis patients compared to non-dialysis individuals. To determine whether this significant difference was attributable to the dialysis procedure itself, sPD-L1 levels were compared before and after the initiation of dialysis. Evaluation revealed an increase of approximately 100 pg/mL in sPD-L1 levels three months after starting dialysis, suggesting that dialysis contributes, at least in part, to the elevation of serum sPD-L1 levels. Additionally, sPD-L1 levels were directly related to dialysis duration. The molecular weight cutoff for molecules filtered by low-flux and high-flux hemodialysis membranes typically ranges from 5 to 15 kDa and 50 to 60 kDa, respectively. Considering its molecular weight, i.e., 37–56 kDa (8), it is likely that sPD-L1 accumulates in the blood, as it is not effectively filtered by low-flux dialysis, which was used in most patients in the Dialysis Initiation and Maintenance cohorts in this study. In patients on dialysis, urine volume progressively decreases as dialysis vintage increases (22). It is hypothesized that initiation of dialysis reduces urine volume, leading to a decrease in the amount of sPD-L1 excreted in urine, and hence, its elevation. Platelet activation (23) and vascular endothelial cell damage or stimulation (8) caused by hemodialysis procedure itself or an arteriovenous fistula might also have contributed to the observed elevation in serum sPD-L1 levels in dialysis patients in this study. Previous studies have suggested that sPD-L1 levels are higher in patients with coronary artery disease than in those without (8), and that elevated sPD-L1 levels, induced during the development of atherosclerosis, are associated with future cardiovascular events (24). These findings suggest that sPD-L1 levels might reflect the severity and instability of atherosclerotic lesions as well.

While CKD typically advances with aging, there are cases where its progression outpaces what would be expected based solely on chronological age. The drivers of this accelerated decline are complex and multifactorial, including cellular senescence, inflammation, mitochondrial dysfunction, disruptions in sirtuin and Klotho signaling, and dysregulation of the autophagy-lysosome system (25). These overlapping mechanisms make it difficult to pinpoint a singular cause. Nevertheless, Pippin et al. demonstrated that PD-1 expression on podocytes increases with age, a phenomenon closely tied to renal functional decline and worsening glomerulosclerosis due to podocyte apoptosis caused by PD-1/PD-L1 pathway upregulation (2). Interestingly, blocking PD-1 with an antibody has been shown to reduce both podocyte senescence and inflammation, thus extending the lifespan of these critical cells (2). In the present study, sPD-L1 levels increased significantly with age, independent of sex and kidney function, and also exhibited significant variability even within the same age and sex group. Contrary to the previous hypothesis, elevated sPD-L1 in blood, beyond what is expected with aging, might stimulate PD-1 on podocytes, thereby impairing renal function and advancing CKD progression. However, PD-L1 overexpression can also occur with aging podocytes, inflamed renal tubular epithelial cells (26), renal interstitial tissues involved in fibrosis development (27), and kidney function decline itself, suggesting that these additional factors should be carefully considered in future clinical investigations. Moreover, a recent meta-analysis of randomized clinical trials demonstrated that combined use of anti-PD-1 or PD-L1 antibodies with chemotherapy increased the risk of severe acute kidney injury and the composite of all severe acute kidney adverse events (28, 29). Therefore, blocking the interaction between PD-L1 and PD-1 with antibody therapy might not necessarily halt age-related CKD progression, and could potentially worsen it, probably through autoimmune mechanisms.

This study has several limitations. First, each cohort was primarily enrolled for different study purposes, resulting in notable differences in baseline characteristics. The Vaccine Cohort consisted of relatively healthy individuals with few comorbidities, the Cancer Cohort included patients with esophageal, gastric, or colorectal cancer, and the Dialysis Initiation and Dialysis Maintenance Cohorts comprised patients who had newly started dialysis or were on stable dialysis, respectively. Second, in the Vaccine and Cancer Cohorts, since creatinine and sPD-L1 levels were measured from the same blood samples, it was not possible to determine whether elevated sPD-L1 levels led to renal functional decline or vice versa. Additionally, since an association does not necessarily imply a causal relationship, while patients with elevated sPD-L1 levels in this study exhibited advanced CKD stages, caution is needed when interpreting these findings. Third, although a significant increase in sPD-L1 levels was observed after the initiation of dialysis, the Dialysis Initiation Cohort included only 15 participants, which was fewer than in other cohorts, raising the possibility that this could be a chance finding. Fourth, since PD-1 on podocytes in the renal glomeruli were not evaluated pathologically, it remains unclear whether the interaction between circulating sPD-L1 and PD-1 on podocytes accelerated kidney aging and contributed to CKD stage progression. Fifth, when measuring sPD-L1 levels across the four cohorts, ELISA kit lot numbers were consistent within each cohort, but differed between cohorts. Hence, it was not possible to make subtle comparisons between different cohorts. For example, while dialysis patients had levels three times higher than non-dialysis patients, with the lot number differences falling within the margin of error, in cases where the differences were smaller, such as between cancer and non-cancer patients, it was necessary to avoid combining patient data across cohorts.

To the best of our knowledge, this is the first study demonstrating the association between serum sPD-L1 levels and CKD progression, with sPD-L1 reaching the highest levels in patients with stage 5 CKD. Serum sPD-L1 levels also increased with dialysis initiation and dialysis duration. Further investigation is needed to understand the mechanism of sPD-L1 elevation and its impact on patients with CKD.