Abstract

Introduction

Metformin is a well-tolerated drug that is commonly used to treat diabetes mellitus and insulin resistance. Several retrospective studies have reported that diabetic patients taking metformin had a lower cancer incidence, prompting evaluations of metformin as an anticancer agent. In 2005, Evans et al. first reported an association between decreased cancer incidence and metformin use [1], while subsequent studies showed that diabetic patients treated with metformin were less likely to be diagnosed with cancer and/or to die from it. Post-hoc analysis of the ZODIAC trial showed that patients with type 2 diabetes are overall at increased risk for cancer mortality, however metformin use was associated with lower cancer mortality (HR: 0.43; 95% CI: 0.23–0.80) [2].

Two-thirds of adult males in the United States are overweight or obese, and several studies have demonstrated an association between obesity and incidence of prostate cancer (PCa) [3]. Obese patients have reduced adiponectin levels and increased leptin levels, which may promote PCa growth [4]. In addition, an increased insulin level in a patient with metabolic syndrome can cause an increase in free androgens by decreasing sex hormone-binding globulins, which may also lead to a worse PCa prognosis [5]. In patients who underwent radical prostatectomy, obesity was associated with more aggressive cancer, a higher rate of PSA failure and a trend toward increased risk of positive surgical margins [6]. A meta-analysis of 16 studies analyzed 26,479 PCa patients after their curative treatment and found that an increase in BMI of 5 kg/m² was associated with a 21% increase in risk of primary treatment failure (HR: 1.21, 95% CI: 1.11–1.31, P < 0.01) [7].

Despite promising preclinical evidence of metformin’s antineoplastic activity, clinical studies of metformin monotherapy in PCa have been somewhat confusing. A phase 2 study enrolled 42 chemotherapy-naïve patients with metastatic castration-resistant prostate cancer (mCRPC) and treated them with metformin 1000 mg twice daily until progression. Metformin showed modest activity: 36% of patients were progression-free at 12 weeks; 9.1% were progression-free at 24 weeks, 2 of whom had PSA declines of ≥ 50% [8]. Another clinical study treated 24 men with newly diagnosed PCa with metformin 500 mg 3 times daily prior to radical prostatectomy for a median of 41 days (range 18–81 days). Compared to baseline biopsy, treatment with metformin reduced the Ki67 proliferation index by 29%, while PSA reduction was not statistically significant (P = 0.08) [9].

Colquhoun and colleagues investigated whether the anticancer activity of metformin could be further improved by combining it with bicalutamide. They randomized LNCaP prostate cancer xenografts on a high-carbohydrate/high-fat diet into 4 treatment groups: control, metformin, bicalutamide, or metformin plus bicalutamide. The study revealed no significant difference in tumor volume in mice treated with metformin monotherapy, while mice treated with the combination therapy had significantly smaller tumors and lower PSA than controls and mice receiving either monotherapy (P < 0.0004). [10].

Based on available preclinical and clinical data, we hypothesized that metformin could enhance the antitumor effects of bicalutamide, effectively delaying the onset of metastases or need for testosterone suppression. Here we present results from the multicenter, open-label, randomized phase 2 study of bicalutamide with or without metformin in overweight or obese patients with BCR.

Materials and Methods

Study design

Patients were randomized 1:2 to Arm A (observation x 2 cycles, followed by bicalutamide 50 mg PO daily x 6 cycles) or Arm B (metformin 1000 mg twice daily x 2 cycles, followed by metformin 1000 mg twice daily plus bicalutamide 50 mg PO daily x 6 cycles) (Figure 1) and stratified according to prior definite therapy (radiation vs. surgery). Blood samples were taken at prespecified time points to assess immune activation at baseline, after 8 weeks, and at the end of the trial (32 weeks). The primary endpoint was the proportion of patients with undetectable PSA (< 0.2 ng/mL) at 32 weeks. Adverse events were assessed by the investigators and graded per the NCI’s Common Terminology Criteria for Adverse Events v.4.0. An early stopping rule for futility was set at 39 patients, but due to slow accrual an unplanned interim analysis was undertaken after 29 patients were enrolled.

Open in a separate window

Figure 1.

Consort flow diagram.

Of 42 patients officially screened for the trial, 29 patients were found eligible and randomized, 28 patients completed all 32 weeks of the trial, and one patient in Arm B was lost to follow-up.

Results

Exploratory Immune analyses

Significant differences were observed when the percent change of specific refined immune subsets was evaluated between the treatment arms. Compared to patients undergoing observation (Arm A), 8 weeks of treatment with metformin alone (Arm B), led to a greater reduction in the frequency of PD-1 expressing NK cells (Figure 2A), and a greater increase in the frequency of Immature NK cells (CD56^brCD16⁻) that express the activating receptor NKG2D (Figure 2B). Compared to Arm A, receiving sequential treatment of metformin followed by metformin plus bicalutamide (Arm B) had greater increases at 32 weeks vs pre in the frequency of double positive (DP, CD56^brCD16⁺) NK cells (Figure 2C), and greater reductions in in the frequency of effector memory (EM, CD45RA⁻CCR7⁻) CD4⁺ T cells, including those that express the immune checkpoint CD73 (Figure 2D, ​,E),E), as well as greater decreases in PD-1 expressing EM CD4⁺ T cells, EM CD8⁺ T cells, and Mature (CD56^dimCD16⁻) NK cells (Figure 2F–H).

Open in a separate window

Figure 2.

Difference in % change in refined PBMC subsets after 8 and 32 weeks of treatment (vs baseline) between Arms.

(A-B) Difference in changes in immune parameters after 8 weeks of therapy between Arm A (Observation) vs Arm B (Metformin). (C-H) Differences in changes in immune parameters after 32 weeks of therapy between Arm A (Observation followed by Bicalutamide) vs Arm B (Metformin followed by Metformin plus Bicalutamide). Graphs display values for individual patients plus median and IQR. DP NK (double positive, CD56brCD16+); EM (Effector Memory, CD45RA- CCR7-); Immature NK (CD56brCD16−); Mature NK (CD56dimCD16+).

Discussion

Several recent studies have reported the positive benefit of metformin in the incidence and prognosis of PCa, making it a promising anticancer agent which has been tested in prospective trials. However, extensive research has failed to identify the exact mechanism of metformin’s anticancer effects. Possible mechanisms include activation of the AMPK signaling pathway [2, 16], inhibition of protein synthesis [17], activation of the tumor suppressor p53 [18], induction of apoptosis and/or cell cycle arrest [18], activation of the immune system (protecting tumor-infiltrating lymphocytes from apoptosis) [19], or killing of cancer stem cells [20]. To the best of our knowledge, this is the first randomized, phase 2 trial to evaluate metformin treatment in men with biochemical recurrence of PCa.

From December 2015 to January 2020, 29 patients were randomized and treated on this 3-center, investigator-initiated phase 2 clinical trial. Although as anticipated, treatment was well tolerated with no dose reductions or discontinuations due to safety, the study was not going to reach its primary endpoint showing a difference in the proportion of patients with undetectable PSA values at the end of study and thus was stopped for poor accrual and futility. Although PSADT is a better surrogate for survival than PSA response alone, limited evidence in specific settings (such as hormone-sensitive disease) suggests that PSA response may be used as a surrogate for an ultimate effect on survival and therefore may be used for early stopping in case of futility [21]. Our study showed that fewer participants had undetectable PSA in the combination arm compared to the bicalutamide-only arm, likely due to chance since study did not have enough power due to small sample size.

However, a key secondary endpoint (number of patients with a PSA decline with metformin monotherapy after 8 weeks) was reached. Eight patients (40%) had a PSA decline ranging from −4% to −24%, suggesting modest anticancer activity. It is important to note that all patients had rapidly rising PSA with a PSADT of 3–9 months at baseline, prior to enrolment.

Three other prospective studies of combination treatments with metformin in mCRPC have been published since 2015, when our study was initiated. The combination of metformin plus abiraterone in mCRPC after PSA progression (abiraterone failure) reported no clinical benefit [22], while the combination of metformin and enzalutamide in mCRPC was reported as active. A PSA response of ≥ 50% was observed in 19/24 patients (79%), although this was not a randomized study and patients had not been exposed to enzalutamide [23]. A third study, a randomized phase 2 trial of docetaxel vs docetaxel plus metformin in mCRCP (NCT01796028), was also reported as negative in 2019 [24].

Our immune correlative analyses suggest that metformin, given alone and/or in combination with bicalutamide, has some modest immune potentiating activity, as evidenced by reductions in the frequency of subsets (NK, CD4⁺ and CD8⁺ T cells) that express the exhaustion marker PD-1, reductions in CD4⁺ T cells that express CD73, which is involved in adenosine metabolism and implicated in suppressing an anti-tumor immune response [25], and increases in NK cells that express the activating receptor NKG2D. Others have similarly reported in lung cancer patients with diabetes, that metformin can reduce T cell exhaustion [26].

While this study was negative, observed PSA responses to metformin monotherapy are interesting and warrant further research. This study also demonstrated that metformin can be safely prescribed to non-diabetic patients at higher doses without significant toxicities and with good compliance. Unanswered questions still remain: 1) Is 1000 mg twice daily sufficient therapeutic dosing of metformin as an anticancer agent? 2) What is the duration of PSA response and what is the resistance mechanism to metformin? 3) What is the efficacy in patients with a BMI < 25? and 4) Would diabetic patients with PCa receive the same or greater benefit from metformin given its ability to reduce insulin resistance, insulin levels, and circulating levels of glucose, which can affect cancer progression?

It is hoped that ongoing trials of metformin treatment for PCa, including the Metformin Active Surveillance Trial (MAST), which is recruiting men with low-risk cancer (NCT01864096), and a large phase 3 trial adding metformin to standard therapy for newly diagnosed metastatic castration-sensitive cancer (STAMPEDE – Arm K; NCT00268476) will shed more light on the role of metformin in the treatment of PCa.

Supplementary Material

Acknowledgments

Footnotes

References