Immune checkpoint inhibitors (ICI) have a wide range of medical applications in various malignancies including lung cancer (1), melanoma (2), renal cell carcinoma (3) and squamous-cell carcinoma of the head and neck (4). Among a large number of antibodies, nivolumab, pembrolizumab, and ipilimumab for combination therapy, represent the most frequently used checkpoint inhibitors. Local tissue cells, activated immune cells, as well as malignant cells are able to express programmed cell death ligand 1 (PD-L1) in order to influence tumor defense (5). Nivolumab and Pembrolizumab are antibodies that inhibit the programmed cell death protein 1 (PD-1) signaling, decrease PD-L1- and PD-L2-mediated immune silencing and ultimately lead to an increased anti-tumoral T cell activity, mainly mediated by cytotoxic T lymphocytes (6). Although ICIs have shown a high clinical efficacy in several studies, not all patients benefit from this new type of immunotherapy (7). The search for biomarkers that can help predict tolerance and response (including its long-time sustainability) to this promising type of therapy is ongoing. Previously, it has been shown that increased intratumoral expression of PD-L1 predicts a better treatment response to ICIs (8). Further studies indicate that tumors with microsatellite instability (MSI) seem to be characterized by superior response to immunotherapy using ICIs (9). Other factors such as the tumoral immune cell infiltration, tumor mutational burden, and some tumor specific mutations [e.g. BRAF-mutant non-small cell lung cancer (NSCLC) (10)], have also been associated with favorable treatment response (11). Nevertheless, predictive biomarkers of response per se as well as its long-term benefits have not yet been identified (11).

Telomeres are repetitive DNA sequences located at the ends of the chromosomes. Intact telomeres are essential for genomic integrity and the maintenance of the cellular proliferation potential (12). During life, telomeres shorten with each cell division and somatic cells become senescent or undergo apoptosis once reaching a critically short TL (13). TL represents an established biomarker for the replicative history in specific tissues, turnover of hematopoietic stem cell populations (14–16) as well as leukocyte subsets (14, 17) and the biological aging process of an individual organism in general (18). For various hematological diseases (13), TL has been shown to predict treatment response, such as in chronic myeloid leukemia (19, 20), or has been demonstrated to correlate with the response to immunosuppressive therapy in aplastic anemia (13, 21). However, it currently remains unknown whether the individual TL plays a role in treatment response and outcome in patients undergoing ICI therapy. Thus, the primary aim of our study was to investigate whether the TLs of peripheral blood cells in subjects undergoing ICI therapy is a possible biomarker for therapy response and overall survival (OS).

Despite of the broad application of ICIs in various tumor entities, few factors have been established predicting treatment response and OS to ICI treatment (30). TL represents an established biomarker in various hematological diseases (13, 19–21). The impact of TL on the survival of patients under ICI therapy has not been investigated before. In our study, we were able to demonstrate that patients with longer age-adapted ΔTLs had an improved OS compared to patients with shorter ΔTL, independent of the underlying malignant disease. Of note, we observed no differences in treatment response to ICI therapy with respect to patients’ ΔTL. Furthermore, we were not able to demonstrate an association between ΔTL and tumor type, stage of disease, age, gender, previous treatments of patients or IRAEs. We could not observe the previously described shorter ΔTLs of smokers compared to non-smokers possibly due to the small sample size (31). However, we were able to confirm previous data showing that an increased body-mass-index (BMI) had a positive impact on OS in patients receiving ICIs (32).

In the present study using univariate and multivariate analyses, we show that the individual ΔTL in our cohort with heterogeneous solid tumors had a significant impact on the patients’ OS. This effect was most significant when using an ideal ΔTL cut-off value of 0.17. At this cut-off value, the median OS in patients with low ΔTL was only 5.7 months compared to 18.0 months in patients showing a ΔTL above this cut-off value. Since ΔTL is not specifically correlating with known factors for survival as e.g., tumor stage, tumor entity, performance score or response to treatment, we explain the differences in the survival curves with the fact that TLs might represent a possible marker for the assessment of biological age. Due to the relatively small subgroup sizes, no tumor entity-specific analyses were performed. Furthermore, we cannot establish a clear association with ICI therapy because we lack a comparison group receiving any other type of treatment, such as chemotherapy.

Different studies have shown in the past that telomeres shorten as part of the physiological aging process and TL of peripheral blood leukocytes represent an established global marker for the aging process of the entire organism (12, 18). Supporting our hypothesis of shorter TL as marker for the older biological age, previous studies demonstrated that within the general population, shortened TLs are associated with increased all-cause mortality (33). In oncological patients, a comprehensive study including over 47,000 cancer patients revealed that TL correlates with the outcome of cancer patients (34). In line with this observation, performance status scales like “ECOG” or the “Karnofsky-Index” as an indirect marker for biological fitness were shown to predict survival in oncological patients (35). Furthermore, it was previously shown that also immune cell ageing based on different markers was associated with increased mortality (36). We therefore speculate that shorter TLs will also be associated with an increased non-tumor related mortality in our collective as a possible cause for observed difference in OS. This effect could be amplified by pre-existing diseases and tumor-related complications. However, we have insufficient information regarding the specific cause of death of our patients or detailed information about patient’s pre-existing diseases to definitively prove our hypothesis.

An open question in the field is whether additional factors can specifically cause TL shortening in peripheral blood leukocytes. A variety of environmental factors or chronic diseases can in addition accelerate the physiological telomere related aging process in leukocytes (37, 38). In addition, tumor proinflammatory milieu and its microenvironment may also contribute to TL shortening. Data suggest that tumor-associated inflammation can lead to age-independent immune senescence with consecutive shortening of the TL (39, 40). Due to the old age of our cohort, congenital defects in telomerase-associated proteins leading to impaired telomere maintenance seem to be unlikely as major factor.

Lymphocytes are crucial cells for tumor defense and are substantially involved in maintaining an anti-tumoral immune response, mainly within the context of ICI therapy (1). It is tempting to speculate that premature aging in the T-cell compartment results in immune cell exhaustion explaining the impaired OS in patients with shorter TL (41). This issue becomes of particular interest since preliminary work has shown that telomere shortening can predominantly occur in specific immune cell subpopulations, e.g., in memory cytotoxic T cells (42). Due to methodological constraints, we only measured ΔTL of all leukocytes, and it is quite possible that important differences in ΔTL that may predict treatment efficacy are only evident in subpopulations, such as lymphocytes. In the case of immune cell exhaustion, we would first expect a shortening of the ΔTL in the lymphocyte compartment. However, it is therefore all the more surprising that we can already detect differences in survival in a small cohort and with a global screening of ΔTL. Although our study did not demonstrate an inferior response to ICI in patients with shorter TL, the significantly shortened survival time indicates that we may not have captured (small) differences in treatment response or IRAEs due the small patient cohort as well as the selection bias of our patient population. It is possible that subtle differences may be hidden by our basked study design.

Our study has some limitations as the analyzed patient cohort is heterogenous including patients with different tumor entities who receive different ICIs. We also did not include patients who were treated with different therapeutic approaches such as chemotherapy, radiotherapy, or surgery. Moreover, validation of our exploratory findings in larger patient cohorts, that we hope to have stimulated with our study, is essential before a potential clinical implementation can be considered.

Together, in a collective of 84 patients who underwent therapy with ICIs, we observed that ΔTL at treatment initialization was an independent predictor of OS. We explain our findings with a hypothetical older biological age represented by a premature TL shortening. The pathophysiological background of our observation appears to be unclear. An impact of TL on therapy response or on therapy-associated side effects could not be proven. Additional studies are needed to analyze if further mechanisms that lead to TL shortening in patients with solid tumors exist and to find out if long TLs, especially in lymphocytes, support immunosurveillance competence under ICI therapy during long-term observations.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by Ethics committee of the University Hospital RWTH Aachen, Germany (approval EK 206/09). The patients/participants provided their written informed consent to participate in this study.

SL, TL, BR, and FB designed the study. JG and SL recruited patients. BR and AA performed experiments. SL, JG, FB, and BR performed statistical analysis and generated Figures and Tables. MT, CR, MV, FT, FH, TB, and TL provided intellectual input. BR, SL, JG, FB, and TL drafted the manuscript. SL, BR, FB, and JG revised the manuscript. All authors contributed to the article and approved the submitted version.

This publication is part of a project that has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program (Grant agreement No. 771083). The lab of TL was further supported by the German Cancer Aid (Deutsche Krebshilfe 110043 and a Mildred-Scheel-Professorship), the German-Research-Foundation (SFB-TRR57/P06, LU 1360/3-1, CRC1380/A01, and CA 830/3-1), the Ernst-Jung-Foundation Hamburg, the IZKF (interdisciplinary centre of clinical research) Aachen and a grant from the medical faculty of the RWTH Aachen. Margherita Vieri receives funding from the “Aachener Krebs- und Leukämiehilfe”. FB was funded by a START Grant (No. 691743, RWTH Aachen University) and part of the work was supported by a grant of the “Württembergischer Krebspreis 2019” to FB.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.