Incidental changes in hemoglobin levels in patients with myelofibrosis receiving treatment with sodium-glucose co-transporter-2 inhibitors

Sodium-glucose co-transporter-2 inhibitors (SGLT-2I) (also known as gliflozins or flozins) are increasingly prescribed for diabetes mellitus, heart failure and chronic kidney disease.¹ Food and Drug Administration (FDA)-approved drugs within the SGLT-2I class, bexagliflozin, canagliflozin, dapagliflozin, empagliflozin, ertugloflozin, and sotagliflozin offer cardio-protection and enhance erythropoiesis through inter-related mechanisms.² The physiology of SGLT2-I-induced erythrocytosis is multifactorial; potential mechanisms include: i) increased erythropoietin (Epo) production; ii) suppression of hepcidin and modulation of iron metabolism; and iii) hemoconcentration.³ In placebo-controlled studies, empagliflozin demonstrated a dose-dependent increase in hemoglobin (Hb) and hematocrit (Hct) levels, with a median increase in Hct of 4.8±5.5% and 5.0±5.3% with empagliflozin 10 mg and 25 mg, respectively, versus 0.9±4.7% with placebo.⁴ SGLT-2I have also been shown to ameliorate anemia in patients with chronic kidney disease; in a post-hoc analysis of the Canagliflozin and Renal Events in Diabetes with Established Nephropathy Clinical Evaluation (CREDENCE)⁵ and Dapagliflozin in Patients with Chronic Kidney Disease (DAPA-CKD) trials,⁶ increase in Hct was 2.4% and 2.3% higher in the canagliflozin and dapagliflozin arms, respectively, compared to placebo. Moreover, resolution of anemia was noted in a higher proportion of patients on dapagliflozin than placebo (53.3% vs. 29.4%). In the light of these observations, we postulate that SGLT-2I might stimulate erythropoiesis and improve anemia in chronic myeloid malignancies such as myelofibrosis (MF).^7-9 Accordingly, in the current study, our primary objective was to evaluate the impact of SGLT-2I therapy on Hb levels and anemia-related outcomes in patients with MF. After approval by the Institutional Review Board for minimal risk research protocol, the Mayo Clinic myeloproliferative neoplasm (MPN) clinical database was queried to identify patients with primary or secondary MF, prescribed SGLT2-I for diabetes mellitus, heart failure or chronic kidney disease, following established diagnosis of MF.¹⁰ The time of SGLT-2I initiation was considered as baseline and for each patient, baseline Hb, changes in Hb, and thrombotic events during SGLT-2I treatment were recorded. Therapeutic interventions for MF including timing of initiation and discontinuation of cytoreductive agents were carefully documented. A total of 16 patients with MF (primary MF [PMF], N=12; post-essential thrombocythemia MF [post-ET MF], N=4) received treatment with empagliflozin (N=10), dapagliflozin (N=4), or canagliflozin (N=2) between July 2019 and March 2024. Median age was 74 years; 56% were males. JAK2V617F mutation was detected in 8 (50%) cases: CALR mutation in 5 (31%), MPL in 2 (13%) cases, while one (6%) case was triple-negative. A history of thrombosis before SGLT-2I use was documented in 8 (50%) patients which included 4 arterial (25%) and 4 venous (25%) events. Seven (44%) patients received cytoreductive therapy, including hydroxyurea (N=2), ruxolitinib (N=4), and momelotinib (N=1); 11 (69%) and 6 (38%) patients were also on antiplatelet therapy and systemic anticoagulation, respectively. Table 1 provides detailed clinical and laboratory characteristics of 16 patients with MF started on SGLT-2I at a median of 12 months (range: 2-73 months) after MF diagnosis. Baseline median Hb value at the time of initiation of SGLT-2I was 11 g/dL (range: 8-15); 6 (38%) patients had Hb <10 g/dL including one case (6%) with transfusion-dependent anemia. At a median treatment duration of 11.5 months (range: 2-73), 2 (13%) patients experienced venous thrombosis while receiving SGLT-2I therapy; both events were chronic pulmonary thromboembolism secondary to pulmonary hypertension. Median follow-up from MF diagnosis and SGLT-2I initiation was four years (range: 1-18 years) and 1.2 years (range: 0.3-6.1 years), respectively, during which 4 deaths and no leukemic transformations were recorded.

Table 2 and Figure 1 highlight Hb changes from baseline to post-treatment Hb levels in all 16 patients with MF receiving SGLT-2I. Fourteen (88%) patients experienced an increase in Hb levels after initiation of SGLT-2I, with an improvement in Hb of ≥1 g/dL in 9 (54%) patients. An increase in Hb of 1.5 g/ dL or greater was documented in 6 (38%) cases (Patients 2, 5, 8, 9, 11, and 15), and 3 (19%) patients displayed an increase in Hb of between 1 and 1.4 g/dL (Patients 4, 7, and 13). Among 6 anemic patients with baseline Hb <10 g/dL (Patients 2, 4, 9, 11, 13, and 14), 3 (50%) showed an improvement in Hb of ≥1.5 g/dL, and 2 (33%) displayed an increase in Hb of 1 g/dL. On the other hand, 2 (12%) patients did not experience any improvements in Hb; Patient 3 had a decrease in Hb and Patient 14 remained transfusion-dependent during SGLT-2I therapy. Details on hematologic parameters and the clinical course of patients with an improvement in Hb of ≥1 g/dL are provided below.

Patient 2, a 77-year-old female with JAK2V167F-mutated post-ET MF with no history of thrombosis and not on cytoreductive therapy, received canagliflozin 100 mg daily for diabetes mellitus and chronic kidney disease. Baseline Hb and Hct were 8.2 g/dL and 27.5%, respectively; one month after initiation of SGLT-2I, Hb and Hct levels peaked at 10.2 g/dL and 33.5%, respectively. Shortly thereafter, two weeks after achieving peak Hb and Hct levels, canagliflozin was discontinued due to pancreatic cysts, and Hb after stopping SGLT-2I was 9.3 g/dL.

Patient 4, a 79-year-old male with JAK2V167F-mutated PMF and a history of venous thrombosis, received empagliflozin 10 mg daily for heart failure with reduced ejection fraction (HFrEF) and diabetes mellitus. Baseline Hb and Hct were 9.4 g/dL and 28%, respectively; four months after initiation of SGLT-2I, Hb and Hct levels peaked at 10.4 g/dL and 31.8%, respectively, and these levels were sustained for 25 months. Patient 5, a 72-year-old male with JAK2V167F-mutated PMF, received dapagliflozin 5 mg daily for diabetes mellitus and chronic kidney disease. Baseline Hb and Hct were 10.9 g/dL and 34.7%, respectively; 15 months after initiation of SGLT-2I, Hb and Hct levels peaked at 13.1 g/dL and 42.2%, respectively, and these levels were sustained for nine months. Dapagliflozin was discontinued due to acute kidney injury, and Hb within one month of stopping the drug was 9.5 g/dL. It is to be noted that he had received intravenous iron two weeks prior to achieving peak Hb and Hct levels. Also, he had been started on ruxolitinib 5 mg twice daily three days prior to peak Hb. Patient 7, a 60-year-old female with JAK2V167F-mutated PMF, on hydroxyurea 1,000 mg daily, received dapagliflozin 10 mg daily for HFrEF. Baseline Hb and Hct were 14.7 g/dL and 45.9%, respectively; one month after initiation of SGLT-2I, Hb and Hct levels peaked at 16.1 g/dL and 49.4%, respectively. Thereafter, eight weeks after achieving peak Hb and Hct levels, dapagliflozin was discontinued, although the reasons for this remain unclear.

Patient 8, a 65-year-old male with MPL-mutated PMF received empagliflozin 25 mg daily for diabetes mellitus. Baseline Hb and Hct were 10.3 g/dL and 30.9%, respectively; six months after initiation of SGLT-2I, Hgb and Hct levels peaked at 11.9 g/dL and 35.2%, respectively, and these levels were sustained for six months. He was started on ruxolitinib 5 mg twice daily ten days prior to peak Hb.

Table 1.Clinical and laboratory characteristics of 16 patients with myelofibrosis at initiation of sodium glucose co-transporter-2 inhibitors.

Table 2.Clinical outcomes of 16 patients with myelofibrosis after initiation of sodium glucose co-transporter-2 inhibitors.

Figure 1.Baseline, hemoglobin (g/dL), post-treatment hemoglobin (%) and change in hemoglobin in 16 patients with myelofibrosis (MF) receiving sodium glucose co-transporter-2 inhibitors (SGLT-2I). DA-a; darbepoetin alfa; HU: hydroxyurea; PT: Patient; MMB: momelotinib; RUX: ruxolitinib.

Patient 9, a 75-year-old male with JAK2V167F-mutated PMF and a history of venous thrombosis before and during SGLT-2I, received empagliflozin 10 mg daily for HFrEF. Baseline Hb and Hct were 8.7 g/dL and 27.7%, respectively; eight months after initiation of SGLT-2I Hb and Hct levels peaked at 11.6 g/ dL and 38.4%, respectively. Notably, ruxolitinib 20 mg twice daily had been discontinued and he had been started on momelotinib 200 mg daily three months prior to peak Hb. Patient 11, a 78-year-old male with JAK2V167F-mutated PMF and a history of arterial thrombosis was started on empagliflozin 10 mg daily for HFrEF. He was also receiving darbepoetin alfa 200 mcg every two weeks. Baseline Hb and Hct were 9.9 g/dL and 33.6%, respectively; six months after initiation of SGLT-2I, Hgb and Hct levels peaked at 11.9 g/dL and 37%, respectively, and these levels were maintained for five months. At that time, empagliflozin was discontinued due to gastrointestinal intolerance, and Hb one month after stopping the drug was 9.9 g/dL.

Patient 13, an 82-year-old female with MPL-mutated post-ET MF received dapagliflozin 5 mg daily for diabetes mellitus.

Baseline Hb and Hct were 9.2 g/dL and 30.6%, respectively; 14 months after initiation of SGLT-2I, Hb and Hct levels peaked at 10.2 g/dL and 31.3%, respectively.

Patient 15, a 75-year-old male with JAK2V167F-mutated PMF and a history of arterial thrombosis, received empagliflozin 12.5 mg daily for diabetes mellitus. Baseline Hb and Hct were 13.7 g/dL and 43.4%, respectively; 28 months after initiation of SGLT-2I Hb and Hct levels peaked at 15.9 g/dL and 48.4%, respectively, and these levels were sustained for 30 months. Thereafter, empagliflozin was discontinued due to gastrointestinal intolerance, and Hb after stopping the drug was 13.5 g/dL.

SGLT-2I use in patients with MPN is understudied; a recent report on 11 patients with ET receiving SGLT-2I therapy showed consistent increments in Hb and Hct levels with a baseline median increase in Hb and Hct of 1.5 g/dL (range: 1-4) and 5.1% (range: 2.6-13.8), respectively.¹¹ A separate study described underlying JAK2-mutated MPN in 9 patients on SGLT-2I and highlighted 4 incidents of thrombotic complications.¹² The current series is the first to underline the erythropoietic activity of SGLT-2I in MF patients and suggests the therapeutic value of SGLT-2I for management of MF-related anemia. Hepcidin is up-regulated in MF, and SGLT-2I-induced suppression of hepcidin and modulation of iron metabolism likely contributes to the observed increase in Hb levels.¹³ The majority of patients experienced an increase in Hb levels after initiation of SGLT-2I with 5 patients reporting sustained Hb improvement of ≥1.5 g/dL for at least 12 weeks, including 2 patients not receiving concurrent therapy; most (80%) of these patients were male and all were JAK2-mutated. Moreover, among 6 patients with baseline anemia (Hb <10 g/dL), 5 (83%) showed improvement in Hb of at least 1 g/dL. With regard to drug safety, 2 venous thrombotic events occurred during SGLT2-1 use; both were unrelated to therapy and none of the patients experienced leukemic transformation. Incidentally, SGLT-2I use was associated with a lower incidence of acute myeloid leukemia in a population-based analysis of patients with diabetes mellitus receiving SGLT-2I (N=718,276) compared to dipeptidyl peptidase 4 inhibitors (N=1,159,112) (Hazard Ratio: 0.67; P<0.01).¹⁴ Furthermore, several preclinical studies have provided evidence for potential anticancer effects of SGLT-2I¹⁵ Taken together, the current study suggests salutary effects of SGLT-2I on Hb levels in patients with MF; however, additional studies are required to better characterize the improvements in Hb level in the context of JAK inhibitors and other cytoreductive therapies.

Footnotes

• Received October 28, 2024
• Accepted November 21, 2024

Correspondence

N. GANGAT - gangat.naseema@mayo.edu

Disclosures

NG has served on the Advisory Board for DISC Medicine and Agios. All of the other authors have no conflicts of interest to disclose.