https://orcid.org/0000-0002-7505-261X
Sodium/glucose co-transporter 2 inhibitors (SGLT2i) slow the progression of chronic kidney disease (CKD) and heart failure without the risk of concomitant hyperkalaemia and with a lower risk of developing acute kidney injury than renin–angiotensin system (RAS) inhibitors.
Apart from their impressive cardio-renal protection, SGLT2i have several pleiotropic effects of interest to nephrologists, especially regarding one of the most common and important complications of CKD, anaemia. Severe anaemia not only increases cardiovascular risk but also negatively impacts the quality of life of patients with CKD.
SGLT2i as a class can increase haemoglobin (Hb)/haematocrit (Hct) levels [1].
The Hb/Hct increase is partially independent of patient anaemic status. For this reason, the additional effect of raising Hb levels should not automatically be seen as a positive, but rather as a double-edged sword, potentially positive in anaemic patients but also potentially negative in non-anaemics, especially those with Hb levels at the upper limits of normality at treatment start.
SGLT2i can raise Hb levels regardless of the underlying disease (people with or without diabetes, chronic heart failure) [2, 3]. However, the mean increase is modest compared with that obtained with erythropoiesis-stimulating agents (ESAs) or hypoxia-inducible factor prolyl-hydroxylase (HIF-PH) inhibitors. Moreover, little is known about the magnitude of the effect in advanced CKD, the patient population usually prescribed ESAs or HIF-PH inhibitors.
We can imagine an additive effect of SGLT2i and ESA/HIF-PH inhibitors on Hb increase and maintenance in patients with advanced CKD (moderate to severe anaemia usually develops starting from stage IIIB–IV). There is little information in this respect since most patients enrolled in the randomized clinical trials testing SGLT2i were not in stage IV or V of CKD and thus were not treated with ESAs. In trials with HIF-PH inhibitors, a small percentage of patients were treated with SGLT2i, as most trials enrolled patients with already advanced CKD. This suggests that the CKD populations receiving SGLT2i and ESAs/HIF-PH inhibitors are somehow distinct. However, the number of patients with advanced CKD receiving SGLT2i is expected to rise in the future.
Theoretically, the additive effect could be beneficial, possibly helping to reduce ESA dose or iron needs. This could be a meaningful advantage, considering that ESA use at high doses has been associated with a higher risk of mortality or cardiovascular events in CKD patients. On the other hand, overshooting should be considered and avoided when adding SGLT2i to those already treated with ESAs and having Hb values within the optimal Hb target range of 10–12 g/dL suggested by the European Renal Best Practice. In this respect, it is worth noting that the Hb target range should be individualized by considering age, physical activity and, most importantly, existing comorbidities and concomitant treatments.
Treating severe anaemia is beneficial for cardiovascular protection, regardless of the drug, as shown by the fact that partial anaemia correction can significantly improve left ventricular hypertrophy. More controversial is whether partial correction of less severe anaemia is beneficial regarding cardiovascular morbidity and mortality. Randomized trials with ESAs consistently report that partial correction of anaemia is associated with lower cardiovascular risks than complete correction. This is particularly true when considering the results of the CHOIR (Correction of Hemoglobin and Outcomes in Renal Insufficiency) trial, with 50% of the subjects having diabetes, and the TREAT (Trial to Reduce Cardiovascular Events With Aranesp Therapy) trial, which was conducted in a CKD population with type 2 diabetes.
Raising Hb levels in heart failure patients with darbepoetin did not also reduce cardiovascular morbidity and mortality. On the contrary, iron treatment significantly improved the outcome of heart failure patients, irrespective of the correction of anaemia. This implies that how anaemia is treated is relevant to achieving cardiovascular protection.
HIF-PH inhibitors were developed to treat anaemia with a completely different mechanism from ESAs. They act by activating the HIF system; among the downstream effects, they increase endogenous erythropoietin production and reduce serum hepcidin. This implies the exposure to much lower erythropoietin levels than with ESA and possibly a better iron availability, especially in inflamed patients.
The results from randomized clinical trials have not lived up to expectations of improved cardiovascular safety compared with ESAs. Today, HIF-PH inhibitors are considered, at best, not inferior to current ESAs in terms of cardiovascular complications, especially in the non-dialysis population.
Considering the significant nephro-cardioprotective effects of SGLT2i, we could speculate that by adding SGLT2i to ESA or HIF-PH inhibitors, we could obtain better control of anaemia with reduced dose needs of ESAs or HIF-PH inhibitors and possibly reduce the risk of cardiovascular complications in high-risk patients.
Partial anaemia correction is the gold standard for treating anaemia in CKD patients. However, because of the decrease in the optimal Hb target range, overall CKD patients are kept at average lower Hb levels than they used to be one or two decades ago. The shift to the left of the curve of distribution of Hb has brought more patients towards lower Hb values. Therapeutic nihilism and ESA hyporesponsiveness have further increased the percentage of patients with Hb levels below the optimal target range. SGLT2i could be then helpful in increasing the number of CKD patients achieving the recommended Hb target range. Moreover, it would be intriguing to speculate that adding SLGT2i to ESA/HIF-PH inhibitors (in the case of low-dose needs) may open the possibility of aiming safely at a higher Hb target range than the one now recommended for ESA/HIF-PH inhibitor therapy alone.
When the rise of Hb levels following SGLT2i therapy was first observed, it was thought to be due to haemoconcentration secondary to the diuretic effect (Fig. 1). This interpretation is partially correct, but it does not explain the maintenance of higher Hb levels in the long term, as the diuretic effect usually vanishes over time. This is especially true in nondiabetic patients and in the advanced stages of CKD, where the diuretic effect is relatively modest in the first weeks of treatments due to a much lower excretion of glucose and sodium because of the absence of diabetes and the reduced nephron number.
Possible mechanisms explaining increased erythropoiesis following therapy with SGLT2 inhibitors. The observed increase in Hb/Hct levels is due to the combination of increased erythropoiesis and haemoconcentration. Increased erythropoiesis can be sustained by increased iron availability, increased production of erythropoietin from the kidney and possibly by the liver. Among possible mechanisms explaining the increased erythropoietin production, either improved hypoxia following reduced oxygen consumption or activation of the HIF system because of local hypoxia could be considered. HIF activation from sirtuin 1 has also been suggested. Finally, decreased inflammation could also have a role in improving the viability of erythropoietin-producing cells in the kidney and thus increase erythropoietin synthesis. EPC, erythropoietin-producing cells; EPO, erythropoietin.
Conversely, SGLT2i increase erythropoietin levels and reticulocyte count. This is coupled with the concomitant change in indices of iron stores, particularly the reduction in plasma ferritin levels. Several mechanisms could explain increased erythropoiesis following SGLT2i therapy (Fig. 1).
First, as SGLT2i block sodium reabsorption in the proximal tubule, renal hypoxia could improve locally as a consequence of reduced oxygen consumption for cellular ion transportation [4]. The improved cellular microenvironment could restore erythropoietin-producing cells’ viability and thus increase their erythropoietin production. Supporting this hypothesis, experimental evidence has shown that SGLT2i therapy suppresses HIF-1α but stimulate HIF-2α, which is the main driver of erythropoietin synthesis. SGLT2i may improve renal oxygenation also through the involvement of vascular endothelial growth factor [5].
As SGLT2i increase sodium delivery to distal tubules, they cause increased workload and oxygen consumption locally and cause a state of ‘physiological hypoxia’ [6], possibly activating several mechanisms or renal protection. However, it is unknown whether this translates into increased erythropoietin synthesis.
Sirtuins have been identified as possible mediators of the mechanism, bypassing the control of prolyl hydroxylase domain (PHD) enzymes. However, data obtained with SGLT2i in nondiabetic kidney disease have shown an opposite effect on the HIF pathway (i.e. the activation of the HIF-1α subunit and the suppression of the HIF-2α one, contradicting the hypothesis) [7].
Other experimental evidence seems to support that SGLT2i may improve erythropoiesis by interfering with proinflammatory pathways [8].
Finally, as for HIF-PH inhibitors, the stimulation of the hepatic synthesis of erythropoietin has been suggested, but there is no experimental evidence in support [9].
Given the multiple mechanisms of action of SGLT2i in raising Hb levels, we wonder whether the Hb increase could be a simple and unifying explanation for their renal–heart protection. If this is the case, the protective mechanism should be similar for the two organs. However, partial or complete anaemia correction with ESA/HIF-PH inhibitors neither slows the progression of CKD nor provides cardioprotection. The diuretic effect could explain the cardioprotective effect, although the cardioprotection associated with the use of diuretics is increasingly associated with a reduction in the sodium pool at the level of the arterial intima and not with the natriuretic effect. Even if the interpretation was correct, no data support it.
Because of their effect on the HIF system, SGLT2i may simulate the action of HIF-PH inhibitors and increase Hb levels. As mentioned above, this is coupled with the concomitant reduction in ferritin levels. Theoretically, the mobilization of iron from depots (possibly associated with improved intestinal absorption) may support erythropoiesis and provide iron to the mitochondria of cardiac cells, improving their function. However, it is unlikely that in the long term, after a progressive depletion of already existing iron stores, the increase in intestinal iron absorption alone could contribute to this hypothetical mechanism of cardio-protection. This is also in light of the fact that there is a progressive reduction in intestinal iron absorption in the population of nephrological interest with the most advanced stages of CKD. Finally, and contrary to expectations, the large phase 3 registration trials of HIF-PH inhibitors have not demonstrated cardiovascular and renal protection compared with ESA or placebo and even raised concerns about cardiovascular safety, particularly in non-dialysis patients [10], the population studied in the trials with SGLT2i.
As an ancillary effect, some HIF-PH inhibitors can decrease total and low-density lipoprotein cholesterol levels, although associated with a reduction in high-density lipoprotein cholesterol. Despite their possible influence on the HIF system, this has not been described with SGLT2i so far. One possible reason for the discrepancy could be that HIF-PH inhibitors act systemically, whereas the effects of SGLT2i regarding hypoxia seem more organ-specific.
To conclude, from a nephrological perspective, SGLT2i added to RAS blockers further and significantly slow the progression of CKD, reducing the risk of reaching end-stage kidney failure; nephroprotection is coupled with concomitant cardioprotection. It is still a matter of debate whether Hb increase could be one of the potential mechanisms and, in our opinion, could be at best considered a contributing factor. Apart from speculative hypotheses, increased erythropoiesis should be considered positively in the majority of patients, especially those with advanced CKD, at least for their theoretical possible sparing effect on ESA use. Caution should be exercised in patients with high Hb levels at treatment start, especially those with polycystic kidney disease.
AUTHORS’ CONTRIBUTIONS
F.L. and L.D.V. equally contributed to the conception, writing and editing of the manuscript.
CONFLICT OF INTEREST STATEMENT
F.L. is or was a member of an advisory board of Amgen, Astellas and Vifor Pharma, and a speaker at meetings supported by Amgen, Astellas and Vifor Pharma; L.D.V. has been member of advisory boards for Astellas, GSK and Travere, and she received speaker fees at meetings indirectly supported by Astellas, Vifor, Bayer, AstraZeneca and Amgen.
Comments