Changes in insulin receptor signaling underlie neoadjuvant metformin administration in breast cancer: a prospective window of opportunity neoadjuvant study

Abstract

Introduction: The antidiabetic drug metformin exhibits potential anticancer properties that are believed to involve both direct (insulin-independent) and indirect (insulin-dependent) actions. Direct effects are linked to activation of AMP-activated protein kinase (AMPK) and an inhibition of mammalian target of rapamycin mTOR signaling, and indirect effects are mediated by reductions in circulating insulin, leading to reduced insulin receptor (IR)-mediated signaling. However, the in vivo impact of metformin on cancer cell signaling and the factors governing sensitivity in patients remain unknown.

Methods: We conducted a neoadjuvant, single-arm, "window of opportunity" trial to examine the clinical and biological effects of metformin on patients with breast cancer. Women with untreated breast cancer who did not have diabetes were given 500 mg of metformin three times daily for ≥2 weeks after diagnostic biopsy until surgery. Fasting blood and tumor samples were collected at diagnosis and surgery. Blood glucose and insulin were assayed to assess the physiologic effects of metformin, and immunohistochemical analysis of tumors was used to characterize cellular markers before and after treatment.

Results: Levels of IR expression decreased significantly in tumors (P = 0.04), as did the phosphorylation status of protein kinase B (PKB)/Akt (S473), extracellular signal-regulated kinase 1/2 (ERK1/2, T202/Y204), AMPK (T172) and acetyl coenzyme A carboxylase (S79) (P = 0.0001, P < 0.0001, P < 0.005 and P = 0.02, respectively). All tumors expressed organic cation transporter 1, with 90% (35 of 39) exhibiting an Allred score of 5 or higher.

Conclusions: Reduced PKB/Akt and ERK1/2 phosphorylation, coupled with decreased insulin and IR levels, suggest insulin-dependent effects are important in the clinical setting. These results are consistent with beneficial anticancer effects of metformin and highlight key factors involved in sensitivity, which could be used to identify patients with breast cancer who may be responsive to metformin-based therapies.

Trial registration: ClinicalTrials.gov identifier: NCT00897884. Registered 8 May 2009.

Figure 1

Proposed mechanism of action of metformin. (A) The potential mechanism(s) of antitumor action of metformin involves both direct (insulin-independent) and indirect (insulin-dependent) actions of the drug. The direct, insulin-independent effects of metformin involve activation of AMP-activated protein kinase (AMPK) in cancer cells via phosphorylation on Thr172 by liver kinase B1 (LKB1) and a subsequent reduction in mammalian target of rapamycin (mTOR) signaling, protein synthesis and cell proliferation. AMPK-independent actions of metformin may also contribute to its anticancer effects. The indirect, insulin-dependent effects are associated with reductions in circulating insulin levels mediated by inhibition of hepatic gluconeogenesis. The resulting decrease in insulin leads to reduced insulin receptor (IR)-mediated cancer cell signaling. (B) The results of the present study indicate that insulin-dependent effects of metformin are important in the clinical setting. Decreases in circulating insulin levels in metformin-treated patients, coupled with reductions in Akt (S473) and extracellular signal-regulated kinase 1/2 (ERK1/2; T202/Y204) phosphorylation in breast cancer cells, suggest reduced IR activation followed by decreases in phosphatidylinositol 3-kinase (PI3K) and Ras-mitogen-activated protein kinase (MAPK) signaling, cell proliferation and survival. MEK1/2, Mitogen-activated protein kinase kinase 1/2; mTORC1, Mammalian target of rapamycin complex 1; OCT1, Organic cation transporter 1.

Figure 2

Changes in insulin receptor expression upon metformin treatment. (A) Allred scores for the insulin receptor (IR) as measured pre- and post-metformin treatment. (B) Representative images are shown for IR pre- and post-metformin treatment with Allred scores of 6 and 2, respectively. Scale bar represents 30 μm.

Figure 3

Changes in phosphorylated Akt and phosphorylated extracellular signal-regulated kinase 1/2 levels upon metformin treatment. Allred scores for (A) phosphorylated Akt (p-Akt) S473 and (B) phosphorylated extracellular signal-regulated kinase 1/2 (p-ERK1/2) T202/Y204 as measured pre- and post-metformin treatment. The overall score (average of nuclear and cytoplasmic scores) for p-Akt is shown. (C) and (D) Immunohistochemical staining of tumor tissue for p-Akt S473 (C) and p-ERK1/2 T202/Y204 (D). Representative images are shown for each phospho-protein pre- and post-metformin treatment. Corresponding Allred scores were as follows: p-Akt S473 pre-metformin = 7, post-metformin = 3; p-ERK1/2 T202/Y204 pre-metformin = 8, post-metformin = 0. Scale bar represents 30 μm.

Figure 4

Organic cation transporter 1 expression in breast tumors. Allred scores for organic cation transporter 1 (OCT1) as measured in the surgical tumor specimen (post-metformin). (A) Graph depicting distribution of OCT1 Allred scores. (B) Immunohistochemical staining of tumor tissue for OCT1. Representative images are shown for strong (Allred 8) and weak (Allred 3) staining. Scale bar represents 30 μm.

Figure 5

Changes in phosphorylated AMP-activated protein kinase and phosphorylated acetyl coenzyme A carboxylase levels upon metformin treatment. Allred scores for (A) phosphorylated AMP-activated protein kinase (p-AMPK) T172 and (B) phosphorylated acetyl coenzyme A carboxylase (p-ACC) S79 as measured pre- and post-metformin treatment. Immunohistochemical staining of tumor tissue for (C) p-AMPK T172 and (D) p-ACC S79. Representative images are shown for each phospho-protein pre- and post-metformin treatment. Corresponding Allred scores were as follows: p-AMPK T172 pre-metformin = 8, post-metformin = 4; p-ACC S79 pre-metformin = 7, post-metformin = 4. Scale bar represents 30 μm.