Transtubular potassium gradient (TTKG)

Clinical background

When a patient presents with a potassium disturbance, the central differential diagnostic question is whether the kidneys are handling potassium appropriately: secreting too much in hypokalaemia, or too little in hyperkalaemia. A spot urine potassium alone is not an adequate measure, since the concentration depends as much on urine flow as on tubular potassium secretion. At a low flow rate the urine potassium may be high even though the kidney is conserving potassium appropriately, and vice versa.

The transtubular potassium gradient (TTKG) was developed to separate potassium secretion in the cortical collecting duct from the effect of flow. The index is semiquantitative and estimates the potassium concentration in tubular fluid at the end of the cortical collecting duct, corrected for water reabsorption distal to that point [1]. It is therefore not a risk stratification instrument but a physiological reasoning tool intended to answer whether the kidney's handling of potassium is appropriate in relation to the serum potassium level.

Calculating the transtubular potassium gradient

TTKG is calculated as:

$$\text{TTKG} = \frac{U\text{-potassium} \times S\text{-osmolality}}{S\text{-potassium} \times U\text{-osmolality}}$$

The variables are urine and serum potassium in mEq/L and urine and serum osmolality in mOsm/kg. The mathematical idea is that the urine osmolality divided by the serum osmolality estimates the degree of water reabsorption in the collecting duct. By multiplying the urine potassium by the ratio of serum to urine osmolality, the urine potassium is "corrected" for this water reabsorption, and the result should reflect the potassium concentration in tubular fluid at the end of the cortical collecting duct.

The derivation study was performed by Ethier and colleagues at the University of Toronto and published in 1990 [1]. The study comprised healthy volunteers in whom hypokalaemia was induced (by potassium depletion of non-renal cause) and hyperkalaemia (by an oral potassium chloride load of 50 mmol). Patients with hyperaldosteronism were also studied, and healthy volunteers were given fludrocortisone to stimulate potassium secretion maximally. The outcome modelled was the TTKG at a known, physiologically defined potassium status.

Interpretation in practice

The TTKG must always be interpreted in relation to the serum potassium: the same absolute value means different things in hypo- and hyperkalaemia. The reference values established by Ethier et al. still form the basis of interpretation [1].

In hypokalaemia of non-renal cause the TTKG was measured at 0.9 ± 0.2, whereas patients with hyperaldosteronism-induced hypokalaemia had a TTKG of 6.7 ± 1.3 [1]. After an oral potassium load in normokalaemic subjects the TTKG rose to 13.1 ± 3.8, and after fludrocortisone stimulation to 11.8 ± 3.6 [1]. This confirms that a kidney actively secreting potassium should have a TTKG above approximately 10 in hyperkalaemia, and that a value below 5 in hyperkalaemia suggests that tubular potassium secretion is inadequate.

Validation and performance

TTKG has not undergone formal external validation in the modern sense with a reported c-statistic or calibration analysis. It is a physiologically derived index, not a prediction model, and its performance has been assessed through clinical cohort studies rather than through measures of discrimination.

Wu and colleagues evaluated the TTKG in a prospective cohort of 99 patients with chronic normotensive hypokalaemia at a tertiary medical centre [2]. A TTKG above 3 was used as the threshold for defining high urinary potassium excretion. This threshold was universally met in patients with renal tubulopathy (Gitelman syndrome, Bartter syndrome, distal renal tubular acidosis), confirming that the TTKG identifies renal potassium loss when it is present. However, more than half of the patients with gastrointestinal causes (anorexia/bulimia, laxative misuse) also had a TTKG above 3, which means that a raised value does not reliably distinguish renal from extrarenal potassium loss [2]. The authors found that the urine sodium-to-chloride ratio was a more discriminating measure in this population.

Limitations

TTKG rests on two physiological assumptions that must be satisfied for the index to be meaningful. First, the calculation presupposes that the urine osmolality reflects the osmolality of tubular fluid at the end of the cortical collecting duct. This holds only when ADH is acting and water reabsorption in the collecting duct is active, which in practice requires the urine osmolality to exceed the serum osmolality. During water diuresis, when the urine osmolality is lower than the serum osmolality, the assumption breaks down and the TTKG becomes uninterpretable. Ethier et al. showed already in the derivation study that the TTKG was low despite hyperaldosteronism during water diuresis [1].

Second, the index presupposes that sodium reabsorption distal to the cortical collecting duct is negligible. If the urine sodium is low (below 25 mEq/L), a substantial proportion of sodium has already been reabsorbed in the collecting duct, and the urine osmolality then no longer reflects conditions at the cortical measurement point. The calculator should therefore be used only when the urine sodium exceeds 25 mEq/L.

A further limitation is that the TTKG does not take account of potassium reabsorption in the medullary collecting duct, which can occur in potassium depletion. Under such circumstances the TTKG may underestimate the actual secretion in the cortical collecting duct. Ethier et al. noted that pre-existing hypokalaemia could lower the TTKG despite high aldosterone activity [1].

In summary, the validity of the TTKG has been questioned in the nephrology literature, and the tool should not be used as the sole basis for decision. It should be interpreted together with the urine sodium, urine chloride, acid–base status and the clinical context. Alternative measures such as the urine potassium-to-creatinine ratio may be considered, particularly when the conditions for the TTKG are not met.

References

1. Ethier JH, Kamel KS, Magner PO, Lemann J Jr, Halperin ML. The transtubular potassium concentration in patients with hypokalemia and hyperkalemia. Am J Kidney Dis. 1990;15(4):309-15. PMID: 2321642
2. Wu KL, Cheng CJ, Sung CC, et al. Identification of the Causes for Chronic Hypokalemia: Importance of Urinary Sodium and Chloride Excretion. Am J Med. 2017;130(7):846-855. PMID: 28213045