Abstract
Background/Aim: Recent evidence suggests potential synergistic antitumor effects of the combination of programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) immune checkpoint inhibitors with the oral hypoglycemic agent metformin. The aim of this study was to investigate the safety and activity of metformin combined with nivolumab in diabetic cancer patients. Patients and Methods: Patients with advanced melanoma, renal cell carcinoma or lung cancer receiving nivolumab with concurrent diabetes treated with metformin were retrospectively collected. The primary endpoint was the safety of nivolumab plus metformin combination. Results: We collected 40 patients with solid tumors who received metformin for concomitant diabetes and nivolumab as anticancer therapy in four Italian Hospitals. The concomitant use of nivolumab and metformin was well tolerated; adverse events (AEs) of any grade occurred in 75% of patients (mainly fatigue, pruritus, rash, and asthenia). Grade 3 AEs occurred only in 20% of cases; no grade 4 AEs were observed. A statistically significant correlation was found between higher doses of metformin (>1,000 mg daily) and longer progression-free survival (p=0.021), overall survival (p=0.037) and higher overall response rate. Conclusion: The combination of nivolumab and metformin was safe and might have an antitumor activity, supporting further investigations on the synergistic antitumor effect of this combination.
In the last years, immunotherapy with programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) immune checkpoint inhibitors (ICIs) has been one of the major innovations in cancer treatment, improving the prognosis of renal cell carcinoma (RCC), melanoma, and lung cancer patients (1-4). However, not all patients benefit from ICIs, encouraging the development of further novel treatment strategies.
One interesting approach under investigation is the possibility of combining an anti-PD-1 antibody (i.e., nivolumab) with the oral hypoglycemic agent metformin, given the crucial role of the cross talk between cancer metabolism and the host immune system in controlling tumor cell growth. The different amount of several metabolic factors in the tumor microenvironment (TME) can elicit both anti-tumor effects on cancer cells and the induction of tumor growth (5). Cancer cells compete with immune cells for the use of glucose by increasing the expression of glycolytic enzymes. Moreover, tumor cells favor glycolysis for energy supply even in the presence of oxygen - Warburg effect -, increasing lactate levels and inhibiting effector T-cell function, which ultimately results in a decreased antitumor immunity (6, 7).
Metformin, a commonly used anti-diabetic drug, which has intensely been investigated for its anticancer properties, is also supposed to elicit the host immune response against tumor cells (8, 9).
Metformin exerts anticancer effects both via an insulin-dependent and insulin-independent way. The first is based on the capacity of increasing glucose uptake by muscle cells, leading to a reduction of blood glucose and insulin levels. High levels of insulin increase the levels of insulin growth factor receptor (IGFR) and, as a consequence, increase the levels of Ras-mitogen-activated proteins leading to cell proliferation. The reduced levels of insulin caused by metformin inhibits anabolic metabolism and enhances catabolic metabolism, ultimately resulting in reduced cell proliferation. The insulin-independent anticancer effects of metformin are based on metformin-mediated inhibition of liver gluconeogenesis by regulating the adenosine monophosphate activated protein kinase (AMPK) and liver kinase B1 (LKB1) pathways. This led to an increased cellular AMP/ATP ratio, and therefore inhibition of mTORC1 through AMPK activation of TSC1/2-dependent or -independent mechanisms. The ultimate result is the inhibition of both protein synthesis and gluconeogenesis (10-12). Other anti-cancer mechanisms of metformin have been investigated, including Rag GTPase-mediated mTORC1 inhibition, attenuation of Ras-induced reactive oxygen species (ROS) production with a corresponding reduction in DNA damage and the formation of mutations, down-regulation of cyclin D1 expression, and arrest of the cell cycle at the G0/G1 phase (13-15).
Regarding the immune-modulating role of metformin, this drug seems to enhance the anticancer immune response by inhibiting of tumor-infiltrating CD8 T cells (CD8 TILs) apoptosis, reducing myeloid-derived suppressor cell (MDSCs) activity through the AMPK-DACH1-CXCL1 axis, and blocking M2-like polarization of tumor associated macrophages (16-18). Moreover, metformin favors the switch from glycolysis to a mitochondrial metabolism by fatty acid oxidation of TILs, thereby potentiating PD-1 blockade and increasing CD8+ T cells anti-tumor response. Metformin synergizes with ICIs also by promoting phosphorylation of PD-L1, its accumulation in the endoplasmic reticulum (ER), and its degradation (5).
Finally, on the other hand, ICIs can have a direct effect on both immune cell and cancer cell metabolism. PD-1 suppresses effector T cell metabolism, including down-regulation of AKT phosphorylation, reduction in glucose and amino acid uptake, and suppression of overall metabolic activity. Anti-PD-1 antibody could lead to impaired mitochondrial metabolic fitness and activation of glycolysis in T cells required for persistent anti-tumor immunity (7, 19, 20).
In conclusion, ICIs are responsible for important metabolic changes, including an important metabolic reprogramming downstream of T cell signaling. Therefore, targeting metabolic parameters during immunotherapy by combining metformin with the anti-PD-1 antibody nivolumab could result in a synergistic antitumor effect. The safety profiles of both drugs (when used separately) are well established, and considering the mechanisms of action of each agent little overlap in their toxicity is expected when used concomitantly. We retrospectively collected data regarding diabetic patients on metformin therapy who were treated with nivolumab for various cancer types, with the aim of investigating the safety of the combination of metformin plus nivolumab as well as the correlation between the dose of metformin therapy and the patients’ outcomes.
Patients and Methods
Patients. Patients of 18 years of age or older who had been diagnosed with advanced melanoma, mRCC, or lung cancer and referred to the Oncology Departments of four Italian tertiary Hospitals, were screened to find those with a known diagnosis of Diabetes Mellitus type II under treatment with metformin and who had received nivolumab, the immune checkpoint inhibitor (ICI) directed against PD-1, concurrent to metformin. Eligible patients were accrued between April 2018 and September 2019. A detailed review of patients’ medical history, including the basic demographics, prior renal function, metformin dose, the start and – eventually – the end of nivolumab treatment, and the best response to ICI therapy were recorded.
Endpoints of the study. The primary endpoint was to assess the safety of the combination of nivolumab plus metformin. The secondary endpoint was to evaluate a correlation between the dose of metformin therapy and the patients’ outcomes in terms of progression-free survival (PFS), overall survival (OS) and overall response rate (ORR).
Adverse events assessment. Drug-related (nivolumab and metformin) adverse events (AEs) were collected for each patient. We selected the most frequent AEs reported with nivolumab in the pivotal trials for each cancer type (AEs of interest), which included nausea, fatigue, pruritus, diarrhea, decreased appetite, rash, cough, anemia, dyspnea, peripheral edema, asthenia, myalgia/arthralgia, pyrexia, pneumonitis, vitiligo, constipation, and impaired renal function. Moreover, any other adverse event that was judged by clinicians to have a causal relationship with the combination of the two drugs of interest (nivolumab plus metformin) was reported. An adverse event (AE) was defined as any new untoward medical occurrence or worsening of a pre-existing medical condition in a patient administered the investigational drugs. An AE could therefore be any unfavorable and unintended sign (including an abnormal laboratory finding, for example), symptom, or disease associated with the use of the investigational drugs, whether or not considered related to the investigational drug. The causal relationship to the study drug was determined by the physicians and defined as related if there was a reasonable causal relationship between the study drug administration and the AE. The frequency and severity of AEs were graded according to NCI CTCAE V4.0.
Statistical analysis. Patients were eligible for analysis if they received metformin therapy and underwent at least one infusion of nivolumab. Descriptive statistics were used to characterize key demographic and clinical variables at baseline (defined as the date of starting nivolumab therapy), and to assess the incidence of drug-related adverse events.
OS was evaluated from the start of nivolumab therapy to death or last follow-up whichever occurred first. PFS was evaluated from the start of nivolumab therapy to disease progression according to RECIST v1.1 criteria. To evaluate ORR, disease assessments were performed every 3 months according to the clinical practice of the enrolling centres. Patients who did not experience an event of interest were censored at the date of last follow-up. The Kaplan– Meier method was used to analyse the time-to-event endpoints of specific survival. Cox proportional-hazard models were used to estimate hazard ratios for OS and PFS. All the variables were significant if p<0.05. The PASW software (Predictive Analytics SoftWare; v 21; IBM SPSS) was used for the analysis.
Results
Study population. From April 2018 until September 2020 a total of 40 cancer patients who fulfilled the inclusion criteria were included in the final analysis. Twenty-four out of 40 patients were lung cancer patients, 13 patients had a clear cell renal cell carcinoma, and the remaining 3 patients had a diagnosis of malignant melanoma. The daily doses of metformin ranged from 500 mg to 3,000 mg. Baseline characteristics of the included patients are reported in the Table I.
Patient characteristics.
Incidence of adverse events of interest. AEs of interest of any grade occurred in 30 of the 40 patients (75%) treated with nivolumab plus metformin (Table II). The most common AEs of interest were fatigue (14 patients, 35%), pruritus (10 patients, 25%), rash (9 patients, 22.5%), asthenia (9 patients, 22.5%), and myalgia/arthralgia (8 patients, 20%). Grade 3 or 4 treatment-related AEs of interest occurred in 8 out of the 40 patients (20%); the most common grade 3 or grade 4 events were fatigue (2 patients, 5%), asthenia (2 patients, 5%), and myalgia/arthralgia (2 patients, 5%).
Incidence of adverse events (AEs) of interest in the study population.
No correlation was found between the metformin dose (≤1,000 mg versus >1,000 mg) and the incidence of AEs of interest. In particular, AEs of any grade occurred in 30 patients; 40% of cases (12 patients) with a metformin dose ≤1,000 mg daily compared to 60% of cases (18 patients) treated with concomitant metformin dose >1,000 mg daily (p=1.0). Similarly, G3-4 AEs occurred in 8 patients, 4 patients (50%) received concomitant metformin ≤1,000 mg daily and 4 patients (50%) treated concomitantly with metformin >1,000 mg daily (p=0.52).
Correlation between metformin dose and patient outcomes. We evaluated the presence of a correlation between the concurrent dose of metformin therapy and patients’ outcomes in terms of both PFS and OS. Among patients receiving concomitantly metformin ≤1,000 mg daily, the median PFS was 2.8 months (95%CI=0.9-4.7) compared to 10.4 months (95%CI=5.3-15.5) of patients treated concomitantly with metformin >1,000 mg daily (p=0.021) (Figure 1A). When the PFS analysis was adjusted for cancer type, the HR for PFS was 0.37 (95%CI=0.17-0.83), p=0.016.
Correlation between metformin dose (≤1,000 mg versus >1,000 mg) and progression-free survival (A) and overall survival (B).
The median OS of patients concomitantly treated with metformin ≤1,000 mg was 8.8 months (95%CI=0.80-16.7) compared to 17.7 months (95%CI=11.8-23.6) of patients concomitantly treated with metformin >1,000 mg daily (p=0.037) (Figure 1B). When the OS analysis was adjusted for cancer type, the HR for OS was 0.46 (95%CI=0.17-1.02), p=0.056.
Regarding the correlation between the dose of metformin and the tumor response, we found that among patients receiving metformin ≤1,000 mg daily, 6.3% achieved a CR or PR, 43.8% had SD, and 50.0% had PD as best response; among patients treated with metformin >1,000 mg daily, 41.7% achieved a CR or PR, 33.3% had SD, and 25.0% had PD as best response (p=0.041) (Figure 2).
Correlation between metformin dose (≤1,000 mg versus >1,000 mg) and overall response rate.
Discussion
The ICI nivolumab showed significant survival benefits for patients with several cancer types; however, its efficacy remains suboptimal, leading clinicians to explore novel approaches. Combining metformin with nivolumab might produce a synergistic antitumor effect based on a mutual influence on tumor cell metabolism, an enhanced commitment of the host immune system against cancer growth, and the intrinsic anti-tumor activity of both molecules. These effects represent the biologic rationale of retrospective clinical studies evaluating the therapeutic interaction of metformin with ICIs and their safety profile (21-24). In particular, these series compared the activity of an ICI alone in nondiabetic cancer patients with the combination of an ICI plus metformin in diabetic cancer patients. The favorable treatment-related outcomes (ORR, PFS, and OS) observed in patients treated with metformin in combination with ICIs did not reach significance, probably, due to small sample sizes of these studies.
We retrospectively collected 40 diabetic patients on metformin therapy who were concomitantly treated with nivolumab as oncologic treatment for RCC, melanoma, or lung tumors with the aim of better delineating the toxicity profile of this combination. As specified in the methods section, we focused our attention on AEs of interest, selecting the most frequent AEs reported with nivolumab in the pivotal trials for each cancer type. In line with the literature, nivolumab plus metformin was well tolerated, with fatigue, cutaneous reactions (pruritus and rash), and asthenia as the most frequent AEs of interest of any grade, occurring in approximately one third - one quarter of patients (Table III). More serious AEs of interest occurred only in 20% of cases, and fatigue, asthenia, and myalgia/arthralgia were the most common ones; of note, no G4 AEs were reported. Moreover, no patients (including those nephrectomized for RCC) had a worsening of the renal function, which was in theory the most feared adverse event of the combination. Finally, the concomitant use of nivolumab and metformin did not result in other adverse events.
Incidence of adverse events (AEs) reported in the phase III trials of nivolumab in RCC, melanoma, and lung cancer tumor types.
To the best of our knowledge this is the first study that found a statistically significant correlation between the dose of metformin administered concomitantly with nivolumab and patient outcomes. We found that patients receiving higher doses of metformin had longer PFS and OS. This benefit persisted even when the analysis was adjusted for the type of cancer. Analogously, the tumor response was significantly higher in the subgroup of patients treated with higher dose of metformin, supporting the antitumor activity of metformin when used together with nivolumab. In line with our results, preclinical data suggest a dose-dependent antitumor activity of metformin: the higher the dose, the more potent is its antitumor activity (25). However, there is no consensus regarding the optimal dose of metformin in non-diabetic patients. We designed a phase II trial that is currently enrolling non-diabetic mRCC patients pre-treated with anti-VEGFR tyrosine kinases inhibitors to receive nivolumab plus metformin as anticancer therapy at the dose of 500 mg twice daily (Eudract CT number: 2019-004916-76) to assess the activity of this combination. Moreover, a phase I study is prospectively testing the safety and activity of nivolumab combined with metformin in patients with refractory/recurrent solid tumors (26). Another phase II study is evaluating nivolumab and metformin in patients with unresectable stage III-IV non-small cell lung cancer unfit for surgery (NCT03048500).
Given the retrospective nature of our study, all analyses are subject to selection biases and not quantifiable imbalance in variables. Multicenter accrual could also have affected the results, considering the heterogeneous physician-dependent reporting of adverse events. Other major limitations include the small sample size, the diverse dose of metformin treatment, and the different cancer types of enrolled patients. These limitations require additional data to validate our findings in larger cohorts. Moreover, the limited number of patients enrolled did not allow assessment of the potential correlation between the concomitant therapy with nivolumab and metformin compared to nivolumab alone and oncologic outcomes (27).
Conclusion
The concomitant use of the anti-PD-1 ICI nivolumab and metformin is well tolerated with no grade 4 AEs. Moreover, a significant correlation between higher doses of metformin and better patient outcomes (longer PFS, OS, higher ORR) supports further investigations for assessing the synergistic antitumor effect of this combination.
Acknowledgements
E.B. is supported by Institutional funds of Università Cattolica del Sacro Cuore (UCSC-project D1-2020/21) and by the Fondazione Associazione Italiana per la Ricerca sul Cancro (AIRC) under Investigator Grant (IG) No. IG20583.
Footnotes
↵* These Authors contributed equally to this study.
↵** These Authors share the last co-authorship.
Authors’ Contributions
Conceptualization: RI, CC, EB, GP. Methodology: EB. CC, GT. Software: CC, RI, SB. Validation: EB, GT. Formal analysis: RI, CC. Resources: SB; FP; SA, FM; MGF; GP; NF; VI; ER; GS; PB; PB. Data Curation: CC, RI, SA. Writing - Original Draft: CC, RI, SB. Writing - Review & Editing: EB, GT. Visualization preparation, creation and/or presentation of the published work, specifically visualization/data presentation: CC, RI. Supervision, Oversight and leadership responsibility for the research activity planning and execution, including mentorship external to the core team: RI, EB, GT. Project administration, Management and coordination responsibility for the research activity planning and execution: CC, RI, SB, EB, GT.
Conflicts of Interest
Roberto Iacovelli: advisory board member for Pfizer, Janssen, Sanofi, IPSEN, MSD, Novartis. Emilio Bria is supported by the Italian Association for Cancer Research AIRC-IG 20583; E Bria was supported by the International Association for Lung Cancer (IASLC), the LILT (Lega Italiana per la Lotta contro i Tumori) and Fondazione Cariverona. E. Bria received speakers’ and travels’ fee from MSD, Astra-Zeneca, Pfizer, Helsinn, Eli-Lilly, BMS, Novartis and Roche. E. Bria received consultant’s fee from Roche, Pfizer. E. Bria received institutional research grants from Astra-Zeneca, Roche. Giampaolo Tortora: advisory board member for BMS and Novartis.
- Received November 29, 2021.
- Revision received January 10, 2022.
- Accepted January 12, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
