Metformin for MPN: teaching an old drug new tricks

In this issue of Blood Advances, Kristensen et al¹ identified an association between metformin use and decreased risk of myeloproliferative neoplasms (MPNs). In this Danish population–based case-control study, 7% of patients with MPN (268 out 3816) had taken metformin compared with 8.2% of the matched general population (1573 out 19 080) without MPNs. Metformin use was associated with lower odds of developing MPNs, with a marked dose-response relationship by cumulative duration in years. Among individuals with long-term metformin use between 5 and 10 years, the adjusted odds ratio was 0.42 (95% confidence interval [CI], 0.29-0.61). This protective effect was observed across all age groups, sex, driver mutations (JAK2-V617F and CALR), and subtypes of classical Philadelphia-chromosome negative MPNs, though most pronounced with polycythemia vera (PV) and essential thrombocythemia (ET). To our knowledge, this study is among the first to examine and report the potential leukemia preventive-impact of metformin.

Philadelphia-negative MPNs comprise a group of chronic leukemias that stem from aberrant clonal expansion of mature myeloid cells. Clinical presentation varies widely across the spectrum of these diseases, but major causes of morbidity and mortality include arterial and venous thromboses, along with transformation to myelofibrosis and acute myeloid leukemia. The majority of MPNs harbor recurrent somatic mutations in JAK2, CALR, or MPL genes, all of which result in the dysregulated activation of the Janus kinase/signal transducer and activator of transcription (JAK/STAT) signaling pathway. The subsequent derangement in immune homeostasis plays a key role in MPN pathogenesis. The mutant hematopoietic clones of MPNs not only thrive in, but also propagate a hyperinflammatory environment through the production of proinflammatory cytokines such as interleukin 6 and tumor necrosis factor alpha among others.² 

Metformin is a synthetic derivative of galegine, a natural product of the plant Galega officinalis (goat’s rue or French lilac), with blood glucose-lowering activity that was first reported in 1957 by the French physician Jean Sterne.³ It is now the most commonly prescribed medication for type 2 diabetes mellitus (T2DM) worldwide. Several epidemiologic studies have revealed decreased solid cancer risk and related mortality among patients taking metformin, but this study augments the findings of a previous retrospective investigation, which reported significantly lower risk for developing hematologic malignancies among veterans taking metformin vs those taking sulfonylureas.⁴ Although the means by which metformin prevents MPNs require further examination to complement the data presented by Kristensen et al, metformin may attenuate leukemogenesis through downregulation of JAK/STAT signaling and subsequent reduction of the inflammatory cytokines that drive MPN. Notable anti-inflammatory mechanisms of metformin on JAK2 V617F-positive MPN cell lines include intracellular reactive oxygen species production and inhibition of downstream mTOR signaling via adenosine monophosphate-activated kinase (AMPK)-dependent pathways, and inhibition of mitochondrial activity and activation of a subfamily of protein tyrosine phosphatase PP2A via AMPK-independent pathways.⁵ 

Chronic inflammation has long been implicated in the development, symptom burden, and progression of myeloid malignancies. Weeks et al found nonmalignant diseases of “inflammaging,” the continual low-grade stimulation of the immune system associated with physiologic aging, are significantly associated with myelodysplastic syndrome (MDS) and chronic myelomonocytic leukemia (CMML).⁶ Surprisingly, the odds of having T2DM were not higher in patients with MDS/CMML, suggesting inflammation is a multifactorial process with distinct relationships with each comorbidity. A major limitation of this study by Kristensen et al is the absence of information on the prevalence and prognostic impact of inflammaging comorbidities on MPNs. Increasing body mass index has been associated with increased risk of MPN development.⁷ Whether the optimization of a patient’s metabolic profile with metformin, or statins as previously reported by the same group of investigators⁸ can prevent MPNs, is an unmet need. Future studies should also consider how lifestyle and social determinants of health impact MPNs, determining whether nontherapeutic, individualized or community-based programs could be implemented alongside therapeutic interventions.

Similarly, critical gaps exist in understanding the disease-modifying properties of metformin for MPNs. MPN evolves from a premalignant state known as clonal hematopoiesis of indeterminate potential (CHIP). The JAK2 V617F mutation is present in about 3% of the general population⁹ and progression from CHIP to overt MPN may ultimately be driven by inflammation.² Interestingly, the risk of cardiovascular (CV) complications is similar whether a patient has CHIP or an overt MPN,¹⁰ indicating a degree of dysregulated immune signaling even in a noncancerous state. Likewise, CV events are the main cause of morbidity and mortality among low-risk patients with MPN, but current guidelines do not sufficiently address how to best modulate CV risk factors for these patients to prevent adverse outcomes. Hence, there may be a role for metformin in patients across the spectrum of MPNs and CHIP given its antileukemic and anti-inflammatory properties in addition to its ability to improve metabolism. Preclinical studies have demonstrated that metformin, with or without ruxolitinib, a selective JAK1/2 inhibitor, can lead to reduction in tumor burden and splenomegaly in JAK2 V617F knockin–induced MPN mouse models.⁵ Taken together with the results presented by Kristensen et al, prospective examination of metformin for cancer prevention and treatment should be pursued. A challenge may be the long-term follow-up needed to assess outcomes for patients with MPN because median survival is well over 10 years for most patients with ET, PV, and lower-risk myelofibrosis. In addition, given the relative rarity of MPNs, the absolute risk reduction may be too small to implement primary prophylaxis.

A strong base of preclinical and epidemiological evidence exists to investigate the antileukemic effects of metformin through prospective clinical trials. Efforts to understand who may derive the most benefit from metformin are critical. Equally important are efforts to identify additional factors that could be modified to reduce the risk of MPN development and progression. Although Kristenson et al report a favorable association between metformin use and significantly reduced risk of MPNs, it remains to be seen if this will translate to repurposing metformin for MPN prevention.

Conflict-of-interest disclosure: G.S.H. receives consultancy from AbbVie, Pfizer, Incyte, Novartis, GlaxoSmithKline, PharmaEssentia, and Bristol Myers Squibb, and reports spouse employment at Regeneron Genetics Center. M.L. declares no competing financial interests.