Janus kinase 2 (JAK2)-unmutated erythrocytosis related to sodium-glucose co-transporter-2 (SGLT-2) inhibitors, leading to increased hematocrit levels and in some cases prompting cessation of gliflozins treatment, is benign and generally should not prompt discontinuation of these drugs in patients with diabetes, heart failure, or chronic kidney disease, according to findings published in the American Journal of Hematology.
“The rise in Hb [hemoglobin]/hematocrit is non-clonal – JAK2 unmutated – or secondary in nature, and therefore not associated with an increased thrombosis risk, which is unlike the case with JAK2-mutated erythrocytosis as in polycythemia vera (PV),” said Naseema Gangat, MBBS, lead author and director of hematology at Mayo Clinic in Rochester, Minnesota.
SGLT-2 inhibitors, including canagliflozin, empagliflozin, dapagliflozin, and ertugliflozin, also known as gliflozins, are commonly prescribed for the beneficial effects on diabetes mellitus, cardiovascular mortality, and other conditions. The drugs increase Hb and hematocrit levels, which are a surrogate marker for cardiorenal protection.
Because of this wide usage, “there has been an unprecedented rise in erythrocytosis referrals to hematology for patients on SGLT-2 inhibitor therapy, which are frequently triggered by [Hb and hematocrit] levels that exceed thresholds for polycythemia,” researchers noted.
However, the outcomes for patients with JAK2-unmutated erythrocytosis, unlike PV, in which JAK2 is mutated, haven’t been well-defined.
Researchers at Mayo Clinic looked at 100 consecutive patients with SGLT-2 inhibitor-associated JAK2-unmutated erythrocytosis; the study is the largest series on the clinical outcomes and management of patients with SGLT-2 inhibitor-associated erythrocytosis.
At the time of erythrocytosis, 62 patients were taking empagliflozin, 21 were taking dapagliflozin, and 17 were taking canagliflozin. Twelve patients had a history of thrombosis. Baseline median Hb and hematocrit levels were 15.5 g/dL and 45.7%, and peak levels were 18 g/dL and 53.3%, with a median increase of 2.5 g/dL and 7.5% at a median of nine months after the start of SGLT-2 inhibitor therapy.
For 29 patients, intervention included phlebotomy or blood donation, and SGLT-2 inhibitors were discontinued in 26 patients, with yeast infection (n=5), erythrocytosis (n=4), poorly controlled diabetes mellitus (n=4), and urinary tract infection (n=3) cited as the most common known reasons.
After a median follow-up of 24 months, 10 thrombotic events were documented: seven arterial and three venous. Researchers found that neither baseline Hb and hematocrit levels nor Hb and hematocrit levels at the time of the event were associated with thrombosis.
Patients who underwent phlebotomy had a higher incidence of thrombosis – six out of 29 patients – than those with or without phlebotomy – four out of 71 patients, researchers found (p=0.03).
The rise in hematocrit with SGLT-2 inhibitors in these patients leads to organ protection through their effects on cardiomyocytes and improves myocardial function, Dr. Gangat said. “As a result, lowering hematocrit through phlebotomy interferes with the aforementioned physiology and unmasks the inherent risk of thrombosis in these patients.”
Dr. Gangat did encourage JAK2 testing to rule out PV but said clinicians should continue gliflozins because of the many benefits. If phlebotomy were to be shown to produce symptom control, it might be an option, but its use should be dictated by symptoms, not hematocrit level.
“One should refrain from treating a laboratory finding and focus on the patient,” Dr. Gangat said. “If [there is] anxiety on the part of the patient or provider, one could consider dose reduction since there was found to be a dose-dependent correlation with peak [Hb and hematocrit] levels.”
Any conflicts of interest declared by the authors can be found in the original article.
Reference
Gangat N, Abdallah M, Szuber N, et al. Sodium-glucose co-transporter-2 inhibitor use and JAK2 unmutated erythrocytosis in 100 consecutive cases. Am J Hematol. 2023;98(7):E165-E167.
