増田 貴博

While sodium-glucose cotransporter 2 (SGLT2) inhibitors exert diuretic effects, they cause only mild changes in body fluid balance, partly through compensatory increases in water intake and vasopressin (AVP) secretion. However, whether SGLT2 inhibitors directly influence central pathways governing these fluid-regulatory responses remains unclear. We here investigated Sglt2 expression in thirst-regulating centers, the organum vasculosum of the lamina terminalis (OVLT) and subfornical organ (SFO) and hypothalamic AVP-synthesizing regions, including the supraoptic nucleus (SON) and paraventricular nucleus (PVN). Droplet digital PCR analysis revealed Sglt2 expression in all regions, at levels comparable to angiotensin receptor 1a, a key regulator of drinking behavior. In situ hybridization confirmed Sglt2 mRNA signals in OVLT, SFO, SON, and PVN. Double in situ hybridization revealed that Sglt2 is co-expressed in approximately 48%-60% of thirst-promoting NOS1 neurons in the OVLT and SFO. Furthermore, Sglt2 was detected in over 80% of AVP-synthesizing neurons and approximately 60% of oxytocin neurons within the SON and PVN. Our findings suggest that SGLT2 inhibitors have the potential to directly modulate central thirst and neuroendocrine responses, offering a novel anatomical framework for understanding their systemic effects on fluid balance and cardio-renal protection.

The combined use of sodium-glucose cotransporter 2 inhibitors (SGLT2i) and loop diuretics may reduce interstitial fluid (IF) retention without excessive plasma volume depletion; however, short-term effects compared with loop diuretic monotherapy remain unclear. This prospective, open-label observational study compared 7-day changes in body fluid status among patients with chronic kidney disease receiving loop diuretics alone (Loop group), SGLT2i alone (SGLT2i group), or SGLT2i plus loop diuretics (SGLT2i+Loop group). Body fluid status was assessed at baseline and day 7. In unadjusted analyses, percent change in estimated plasma volume (ePV) differed among groups (analysis of variance, p = 0.0076) and was lower in the SGLT2i+Loop group than in the Loop group (mean difference, -3.56%; 95% confidence interval, -6.22 to -0.90; p = 0.0059). However, this difference was not significant after adjustment for baseline imbalances and loop diuretic use. Percent change in IF differed overall (p = 0.0030) but did not differ between the SGLT2i+Loop and Loop groups in unadjusted or adjusted analyses. In group-specific analyses, baseline ePV was inversely correlated with percent change in ePV in the SGLT2i+Loop group (r = -0.4185, p = 0.0418), but exploratory interaction analyses showed no significant treatment-group difference in this association. Thus, early ePV and IF responses were comparable between loop diuretic monotherapy and combination therapy after adjustment. This finding supports the hypothesis that baseline intravascular volume status may be linked to early plasma volume responses to combination therapy, a relationship that warrants testing in larger prospective studies.