Authors' Reply: Acute Cisplatin Nephrotoxicity, Sodium-Glucose Cotransporter 2 Inhibitors, and Benefits Beyond Magnesium

We thank Dr. Shah for his interest1 in our JASN article2 and his thoughtful comments. We acknowledge that our study did not clarify how sodium-glucose cotransporter 2 inhibitors (SGLT2i) affect Na+-K+-Cl− cotransporter 2 activity in the thick ascending limb, which warrants further investigation. In the distal convoluted tubule (DCT), the calcium-sensing receptor (CaSR) may sense luminal glucose and activate the With-No-Lysine[K] kinase 4-STE20/SPS1-related proline/alanine-rich kinase pathway, promoting sodium-chloride cotransporter (NCC) phosphorylation3 and possibly NCC stability, potentially contributing to the increased NCC abundance and DCT area observed in SGLT2i-treated models.1 Studies using DCT-specific CaSR knockout mice and ex vivo approaches with pharmacologic agents targeting CaSR and SGLT2 may be instrumental in exploring this mechanism. Dr. Shah suggested that the benefits of SGLT2i in our study may stem from a preventive effect on cisplatin-induced AKI, citing findings by Song et al.4 However, Song et al. used a single high-dose cisplatin injection (30 mg/kg), resulting in severe kidney injury (serum creatinine: 2.25 mg/dl; BUN: 171 mg/dl), with hyaline casts, vacuolization, and tubular lysis. By contrast, we used repeated low-dose cisplatin (2.5 mg/kg weekly for 5 weeks) to minimize injury. We found a modest, nonsignificant rise in serum creatinine (0.50±0.03 versus 0.43±0.01 mg/dl in controls), lower BUN levels (96±4 mg/dl) than Song et al., and no histologic evidence of severe kidney damage. Although empagliflozin modestly improved GFR and BUN, it fully normalized serum magnesium and restored DCT area,2 supporting a direct role in tubular magnesium handling. Notably, in healthy individuals, SGLT2i administration increased phosphorylated NCC in urinary extracellular vesicles, indicating effects on the DCT that occur independently of kidney injury.3 Another distinction lies in the timing of SGLT2i administration. In Song et al.,4 canagliflozin was administered 7 days before and 1 day after cisplatin treatment, consistent with a preventive strategy. In our study, empagliflozin was introduced 3 weeks after cisplatin once hypomagnesemia was established, highlighting a corrective rather than a preventive mechanism. Although we agree that our model reflects acute rather than refractory hypomagnesemia, complete recovery after cisplatin cessation cannot be confirmed. As briefly described in our article, a pilot experiment showed that serum magnesium levels had nearly returned to control 2 weeks after completing a 3-week cisplatin regimen. However, a small but statistically significant difference persisted (approximately 0.3 mg/dl, P < 0.001 versus control). Similarly, Magil et al.,5 using the same cisplatin dose for 3 weeks, reported a sustained reduction in serum magnesium lasting 24 weeks despite transient GFR decline that resolved by week 7. Collectively, these findings support our hypothesis that functional remodeling of the DCT, and possibly the thick ascending limb, underlies the correction of hypomagnesemia by empagliflozin. We appreciate the opportunity to clarify these points. Although it remains unclear whether the same mechanisms apply to individuals with diabetes or other causes, our study provides mechanistic insight into how SGLT2i may reverse tubular magnesium wasting. Further mechanistic studies and clinical trials are needed to define the extent and broader applicability of the antimagnesiuric effects of SGLT2i.