Compensatory Distal Reabsorption Drives Diuretic Resistance in Human Heart Failure

Abstract

Understanding the tubular location of diuretic resistance (DR) in heart failure (HF) is critical to developing targeted treatment strategies. Rodents chronically administered loop diuretics develop DR due to compensatory distal tubular sodium reabsorption, but whether this translates to human DR is unknown. We studied consecutive patients with HF (n=128) receiving treatment with loop diuretics at the Yale Transitional Care Center. We measured the fractional excretion of lithium (FELi), the gold standard for in vivo assessment of proximal tubular and loop of Henle sodium handling, to assess sodium exit after loop diuretic administration and FENa to assess the net sodium excreted into the urine. The mean±SD prediuretic FELi was 16.2%±9.5%, similar to that in a control cohort without HF not receiving diuretics (n=52; 16.6%±9.2%; P=0.82). Administration of a median of 160 (interquartile range, 40-270) mg intravenous furosemide equivalents increased FELi by 12.6%±10.8% (P<0.001) but increased FENa by only 4.8%±3.3%. Thus, only 34% (interquartile range, 15.6%-75.7%) of the estimated diuretic-induced sodium release did not undergo distal reabsorption. After controlling for urine diuretic levels, the increase in FELi explained only 6.4% of the increase in FENa (P=0.002). These data suggest that administration of high-dose loop diuretics to patients with HF yields meaningful increases in sodium exit from the proximal tubule/loop of Henle. However, little of this sodium seems to reach the urine, consistent with findings from animal models that indicate that distal tubular compensatory sodium reabsorption is a primary driver of DR.

Keywords: heart failure; lithium; loop diuretic; proximal tubular sodium.

Figure 1.

Schematic of the renal tubular mechanisms for DR and how the FELi can be used to probe these mechanisms. The mechanism underlying a poor loop diuretic response can be grouped into proximal tubular (Mechanism #1) or distal tubular (Mechanism #2). Lithium reabsorption occurs in parallel with sodium in the proximal tubule and loop of Henle, but is relatively uncoupled to sodium reabsorption in the distal tubule. Thus by measuring loop diuretic induced changes in lithium excretion, we can query the response of the proximal tubule and loop of Henle to the loop diuretic, regardless if the sodium is ultimately reabsorbed in the distal tubule. Li, lithium; Na, sodium.

Figure 2.

Prediuretic FELi was not different between HF patients and non-HF control patients.

Figure 3.

Urine loop diuretic concentration and increase in FELi are related only at low diuretic concentrations. Panel A, fractional polynomial plot of urinary loop diuretic concentration versus the diuretic induced change in FELi. Panel B, C, D are scatter plots of the first, second, and third tertiles of diuretic concentration versus change in FELi. Urine diuretic is expressed as nanograms of furosemide equivalents per mg of urine creatinine. Panel D truncated at 250 ng/mg to demonstrate the central tendency of the data, the full scale figure can be found in the Supplemental Figure 1. Relationship between urine diuretic concentration and the increase in FELi in (A) the overall cohort and (B–D) tertiles of urine diuretic concentration.

Figure 4.

Relative contribution of different intrarenal mechanisms to diuretic-induced increase in FENa. Compensatory distal tubular sodium reabsorption makes the largest relative contribution to diuretic induced increase in FENa. Of the proximal tubular contribution to diuretic response, the major factor driving poor response was inadequate diuretic delivery rather than primary failure at the loop of Henle.

Figure 5.

Regional sodium handling, urine diuretic levels, and plasma aldosterone in groups with high or low loop diuretic–induced increase in FELi or FENa. Group A had a robust natriuretic response at the proximal tubule, but the majority of sodium was reabsorbed in the distal nephron leading to limited natriuresis. Group B had a poor response at the proximal tubule and thus ultimately a poor natriuretic response. Group C had a robust proximal tubular response, a large but proportionately smaller distal tubular compensation than group A, and thus a good natriuretic response. Group D had a modest proximal response, but due to very little distal reabsorption, this group had a good natriuretic response. Notably, group D had significantly lower plasma aldosterone levels then the remainder of the cohort (P=0.001). Values for ↑ or ↓ represent above or below the median value. FELi was used to estimate the amount of sodium exiting the proximal tubule, FENa the amount of sodium excreted in the urine, and the difference in the two parameters the sodium reabsorbed in the distal tubule.