Metformin may inhibit tumor growth in patients with differentiated thyroid cancer (DTC), according to a study in the Journal of Clinical Endocrinology & Metabolism (2013;98:3269–3279).
“This is one of many protocols globally examining the effect of metformin as a treatment for active cancer,” said Kenneth Burman, MD, a coauthor of the study and Director of Endocrinology at MedStar Washington Hospital Center in Washington, D.C. Researchers are currently investigating the drug in many types of malignancy, including those of the breast, colon, pancreas, and prostate. Burman said that to his knowledge, the study is the first to evaluate the clinical aspects and basic pathophysiology of metformin in DTC.
“It's a very a provocative and important study that suggests that metformin may have some beneficial action in thyroid cancer,” commented Michael Pollak, MD, Director of the Division of Cancer Prevention, Professor and the Alexander-Goldfarb Research Chair in Medical Oncology at McGill University and Director of the Cancer Prevention Centre at Segal Cancer Centre at Jewish General Hospital in Montreal. “Although the study is not definitive because it is retrospective, the data justify further research.”
KENNETH BURMAN, MD
The researchers retrospectively analyzed the medical records of patients diagnosed with DTC and type 2 diabetes and then treated or monitored between 1996 and 2011. The study population was divided into diabetics treated with metformin (34 patients) and those not treated with the drug (21 patients), as well as a control group of 185 nondiabetic individuals matched for age, sex, and body mass index.
The size of the tumors was significantly smaller in the metformin group versus the non-metformin and control groups (1.37 ± 0.97 vs. 2.44 ± 1.49 vs. 2.39 ± 1.73 cm, respectively).
The rate of complete remission (CR) was significantly higher in patients taking metformin compared with the non-metformin group (82.4% vs. 57.1%) and comparable with the control group (81.6%). Additionally, there was a significant association between the CR rate and treatment with metformin and the metformin dose, but there was no association with the duration of drug treatment.
A multivariate analysis indicated that extra-thyroidal extension, distant metastases, and lack of treatment with metformin in diabetics decreased the likelihood of achieving a complete remission.
JOANNA KLUBO-GWIEZDZINSKA, MD, PHD
Further analysis showed that age, locoregional metastases, distant metastases, and lack of treatment with metformin in diabetic patients are associated with an increased risk for shortened progression-free survival.
The team also conducted an in-vitro analysis, which suggested that p70S6K/pS6 may be a molecular target of metformin in DTC cells.
Doesn't Show Causation
Asked for her opinion, Naifa Busaidy, MD, Associate Professor of Endocrine Neoplasia and Hormonal Disorders at the University of Texas MD Anderson Cancer Center, said that although overall the study was well done, it was retrospective and therefore demonstrates only association, not causation. The research indicates that metformin may be acting to keep the cancer controlled, she said. For example, the drug may be helping the patient better respond to radioactive iodine (RAI) treatment or to thyroid-stimulating hormone (TSH) suppression.
While the researchers evaluated metformin dose, as groups were subdivided for analysis the numbers got increasingly smaller, she pointed out, adding, though, that the dose that appeared to have an impact on cancer was the dose generally given to diabetics.
Also of note is that the metformin group had smaller cancers than the other two groups did, said Dennis H. Kraus, MD, Director of the Center for Head & Neck Oncology at New York Head & Neck Institute of North Shore—LIJ Cancer Institute. “This suggests that the drug doesn't prevent cancer but may keep it from growing.”
Mechanisms of Action
Direct and indirect mechanisms of action explain why metformin helps to generate beneficial tumor responses in patients with thyroid cancer, said the study's first author, Joanna Klubo-Gwiezdzinska, MD, PhD, Senior Resident at MedStar Washington Hospital Center.
NAIFA BUSAIDY, MD
“By indirect mechanisms, we mean those that inhibit glucose production in the liver, causing the level of insulin to decline,” she explained. The antineoplastic effects are specifically evident in patients characterized by baseline hyperinsulinemia, such as type 2 diabetics or obese individuals. “This mechanism of action might be important, especially in the subset of cancers that thrive in hyperinsulinemic and hyperglycemic environments.”
Pollak speculated that one way metformin might work in DTC patients is through host effects, where the drug doesn't directly impact the cancer but changes the host's hormonal environment to reduce tumor proliferation (as noted in his recent review of the literature: J Clin Invest 2013;123:3693–3700).
Although in some patients taking metformin, insulin levels are reduced, possibly inhibiting growth of the subset of cancers responsive to the hormone, the degree of decline is not large in magnitude. So, while this is a theory for a mechanism of action, “it's not proven,” he said.
Burman noted that metformin may also decrease pituitary TSH levels, a known growth factor of thyroid cancer cells, representing another possible mechanism by which metformin exerts beneficial effects in patients with thyroid cancer.
Metformin's direct mechanism of action appears to be via inhibition of mitochondrial energy production in cancer cells, resulting in decreased adenosine triphosphate (ATP) production and energetic stress, Klubo-Gwiezdzinska said. The change in ATP-to-AMP ratio within the cell activates AMP kinase, leading to downregulation of the mTOR/p70S6K-signaling pathway responsible for cell division, thus downregulating this slowing tumor growth.
Pollak noted though, that although this process has been demonstrated in laboratory studies, no one has shown that metformin at conventional diabetic doses achieves drug concentrations in tumor tissue sufficient to achieve this clinically, although it has been shown in animal models.
Klubo-Gwiezdzinska said that the researchers also found that consistent with the study's in vitro findings, tumors derived from diabetic thyroid cancer patients treated with metformin are characterized by a significantly lower p70S6K level than tissue samples derived from diabetic DTC patients not treated with the drug.
MICHAEL POLLAK, MD
More Treatments Welcome
Although the majority of patients with DTC do well clinically with standard treatment, about 20 to 30 percent will have recurrent cervical lymph node involvement, Burman said, citing J Clin Endocrinol Metab 2012;97:2706–2713. Additionally, the incidence of thyroid cancer is increasing at the fastest rate of any cancer, according to data from the National Cancer Institute's Surveillance Epidemiology and End Results database.
Generally, men older than 45 tend to have more aggressive disease, he added. “Therefore, although a majority of patients do well, a certain minority have progressive or recurrent disease and require further evaluation and treatment.”
In patients who have distant metastases or recurrent disease, metformin could theoretically help cytotoxic therapies work better, Busaidy said. Metformin may also have applications in the majority of DTC patients who do well with standard treatment. For example, the drug could be used as a neoadjuvant to decrease tumor size before surgery or help patients better tolerate RAI or even decrease the need for it.
Oncologists are increasingly selecting which patients should be treated with RAI because the therapy may not be effective in low-risk populations and has side effects, Kraus said. What many oncologists who treat thyroid cancer want to know is what additional agent—whether metformin or another drug—will help with RAI uptake in high-risk patients.
“Metformin is an unusual drug administratively because it is off-patent and drug companies are not investigating it as they would a new molecule,” said Pollak. “If it had just been discovered and it were a proprietary molecule, you would see a lot of excitement and Phase I dose-finding studies.” Instead, clinical trials are asking whether an anti-diabetic dose is antineoplastic, which is a good question but only “part of the bigger issue.”
Klubo-Gwiezdzinska noted that according to ClinicalTrials.gov, about 183 studies are currently being conducted on metformin in different kinds of cancer: “It's a very hot topic,” and while pharmaceutical funding may be minimal, other sources of funding—for example, through the NIH—are available.
A particularly welcome addition to the literature would be prospective, randomized trials of metformin in the neoadjuvant setting or as combination therapy for metastatic DTC patients, Busaidy said.
Investigators are currently trying to develop new metformin analogues in which pharmacokinetic properties allow the drug to accumulate in tumors, Pollak said. “We really don't know enough about the pharmacokinetics and drug distribution to design the optimal clinical trial, but a first step is obvious and in progress in dozens of clinical trials designed to check on the effect of conventional anti-diabetic doses of metformin.”
Klubo-Gwiezdzinska noted that her current observational, retrospective study will form the basis for further prospective clinical trials, and that animal models will be used to verify some of the mechanisms of action associated with metformin.
DENNIS H. KRAUS, MD
Because metformin is known to be generally safe, a prospective, double-blind controlled study to evaluate the drug alone or as an adjunct in patients with thyroid cancer is also feasible, Burman concluded.