Reduced Insulin Sensitivity in Patients with Myeloid Malignancies and Clonal Hematopoiesis Mutations

Cancer survivors are at increased risk for insulin resistance, which can lead to diabetes mellitus, dyslipidemia, and cardiovascular disease. Whether susceptibility to insulin resistance is due to shared onco-metabolic risk factors, like obesity and older age, or if it originates from malignancy and its treatment is unknown. We investigated metabolic and malignancy-associated risk factors and their effect on insulin sensitivity in patients with treated hematological malignancies. Due to previous glucocorticoid exposure, we hypothesized that patients with lymphoid malignancies would be less insulin sensitive.

Nineteen adult patients with treated hematological malignancies but without diabetes mellitus were evaluated with a 2-hour, 75-gram, oral glucose tolerance test (OGTT) and a hyperinsulinemic-euglycemic clamp. Hyperinsulinemic-euglycemic clamps were performed to measure steady-state glucose infusion rate (M-value) as an indicator of whole-body glucose utilization during high-dose insulin stimulation. Decreasing M-values reflect increasing insulin resistance. Patient characteristics were analyzed and compared with OGTT and clamp results.

The majority of individuals had myeloid malignancies [MDS/AML (n=7) and CML/MPN (n = 4)]. The remainder were diagnosed with NHL (n=7) and CML lymphoid blast crisis (n=1). All patients were treated before metabolic testing. Sixteen individuals (84%) received cytotoxic chemotherapy and 3 (16%) were administered only tyrosine kinase inhibitors. Median lines of systemic treatment were 2 (range, 1-6), and median time from cancer diagnosis to metabolic testing was 10 months (range, 3-132). Immediately before OGTTs and clamps, 10 (53%) patients were in a complete response, and 9 (47%) had persistent/stable disease. During OGTTs, impaired fasting glucose (plasma glucose 100-125 mg/dL) and impaired glucose tolerance (2-hour plasma glucose 140-199 mg/dL) were respectively observed in 5 (26%) and 6 (32%) patients. M-values were decreased in patients with impaired glucose tolerance. Overweight and obese cancer patients also had lower M-values than individuals with normal body mass indices (BMI) (Table 1). There were otherwise no differences in insulin resistance per age, sex, performance status, ethnicity, or diabetes family history. Differences in insulin resistance were not observed based on exposure to cytotoxic chemotherapy, non-cranial irradiation, numbers of lines of treatment, response, or malignancy duration.

Previous glucocorticoid exposure did not impact insulin sensitivity. Surprisingly, patients with myeloid malignancies demonstrated lower M-values (6.92 + 0.54 vs. 12.11 + 1.89; p=0.026) than did those with lymphoid cancers. Acquired mutations involved with myeloid malignancies and clonal hematopoiesis (CH) can increase the risk for atherosclerosis and cardiovascular disease, which are sequelae of insulin resistance. Accordingly, insulin resistance was increased in patients with both a myeloid malignancy and mutations involving either DNMT3A, TET2, ASXL1, or JAK2 compared to myeloid malignancies with unknown/absent CH mutations or lymphoid cancers (6.71 + 0.77 vs. 7.28 + 0.72 vs. 12.11 + 1.89; p=0.031). After adjustment for interaction effect in linear regression modeling, both myeloid malignancy and elevated BMI continued to predict lower insulin sensitivity (Table 2).

Pre-diabetes was diagnosed by fasting glucose or OGTT in 26-32% of patients with treated hematological cancers, more than double the prevalence of impaired fasting glucose or impaired glucose tolerance reported by the National Health and Nutrition Examination Survey III. Contrary to our hypothesis, lymphoid cancers and short-term steroids did not contribute to insulin resistance. Instead, myeloid malignancies and CH mutations were associated with reduced insulin sensitivity. Interventions targeting energy balance and obesity will be needed to prevent the harmful sequelae of insulin resistance in long-term cancer survivors. Alternatively, elimination of CH mutations could prove to be a novel treatment of insulin resistance.