BACKGROUND: Preserved function of the renal collecting duct may be essential for response to the vasopressin V2receptor antagonist, tolvaptan (TLV), but the predictors of response to TLV are unknown. METHODS AND RESULTS: Sixty consecutive patients with stage D decompensated heart failure (HF) who had received TLV on a de novo basis were retrospectively enrolled (TLV(+) group). Among them, 41 patients were responders defined according to urine volume (UV) increase after TLV initiation. In the UV-defined responders, plasma arginine vasopressin (P-AVP) had a close correlation with urine aquaporin-2 (U-AQP2; 5.42±3.54 ng/ml; r=0.843, P<0.001). In contrast, 19 were UV-defined non-responders, and they had extremely low U-AQP2 (0.76±0.59 ng/ml, P<0.001 vs. responders) regardless of P-AVP level. On receiver operating characteristic analysis, U-AQP2/P-AVP ≥0.5×10(3)clearly separated the UV-defined responders from the non-responders. We then identified AQP-defined responders as having U-AQP2/P-AVP ≥0.5×10(3). Sixty propensity score-matched HF patients without TLV treatment were examined, and exactly the same number of patients as that of the AQP-defined responders (n=41) was selected. These patients had a poorer survival without TLV than the TLV-treated responders during a 2-year observation period (73.8% vs. 94.8%, P=0.034). CONCLUSIONS: U-AQP2/P-AVP is a novel predictor of response to TLV in patients with decompensated HF. AQP-defined responders may have a better prognosis on TLV treatment.
BACKGROUND: Preserved function of the renal collecting duct may be essential for response to the vasopressin V2receptor antagonist, tolvaptan (TLV), but the predictors of response to TLV are unknown. METHODS AND RESULTS: Sixty consecutive patients with stage D decompensated heart failure (HF) who had received TLV on a de novo basis were retrospectively enrolled (TLV(+) group). Among them, 41 patients were responders defined according to urine volume (UV) increase after TLV initiation. In the UV-defined responders, plasma arginine vasopressin (P-AVP) had a close correlation with urine aquaporin-2 (U-AQP2; 5.42±3.54 ng/ml; r=0.843, P<0.001). In contrast, 19 were UV-defined non-responders, and they had extremely low U-AQP2 (0.76±0.59 ng/ml, P<0.001 vs. responders) regardless of P-AVP level. On receiver operating characteristic analysis, U-AQP2/P-AVP ≥0.5×10(3)clearly separated the UV-defined responders from the non-responders. We then identified AQP-defined responders as having U-AQP2/P-AVP ≥0.5×10(3). Sixty propensity score-matched HF patients without TLV treatment were examined, and exactly the same number of patients as that of the AQP-defined responders (n=41) was selected. These patients had a poorer survival without TLV than the TLV-treated responders during a 2-year observation period (73.8% vs. 94.8%, P=0.034). CONCLUSIONS: U-AQP2/P-AVP is a novel predictor of response to TLV in patients with decompensated HF. AQP-defined responders may have a better prognosis on TLV treatment.