Emanuel Revici used serum potassium and total blood potassium as metrics in Research in Physiopathology as Basis of Guided Chemotherapy: With Special Application to Cancer.

Here we investigate the possibility of using intracellular and extracellular potassium (and metrics computed from them) as biomarkers. The approach discussed is based on Revici’s work, but is somewhat different.

This approach uses serum potassium and red blood cell potassium to calculate total body potassium status and estimated membrane potential using the Nernst equation for potassium.

A case study of one patient using these biomarkers appears below. We are seeking additional patient data for further explanation of membrane potential as a clinically applicable biomarker.

1 Revici’s Interpretation of Potassium

Revici classifies potassium as a cellular level element. As such he looked at the relative values at the cellular level (ICF) compared to the tissue level above (ECF).

The serum potassium measurement is a standard part of a blood panel. To measure the ICF Revici used a total blood potassium measurement described in Chapter 4, Note 8. Given that potassium is primarily present inside cells (~98%) the total blood value can be used as a proxy for the RBC potassium value (as measured by current lab tests). In other words 38 / 0.42 = 90.5. The only issue is they have different values. Here is an (unvalidated) attempt to arrive at a conversion factor.

Assuming the influence of the serum K on the total blood K is neglible we can estimate the intracellular K from \(K_{TB} = K_{RBC} \cdot Hematocrit\) giving us \(K_{RBC} = K_{TB} \div Hematocrit\) for a typical RBC value of 90.5 corresponding to Revici’s typical total blood value of 38 (given an estimated hematocrit of 42%). This corresponds well with the \(K_{RBC}\) normal values seen (~90).

Figure 127 captures Revici’s interpretation of the blood potassium values. The serum and total blood potassium levels indicate both relative excess/deficiency and anaerobic/dysaerobic status.

Serum vs Total Blood K

Serum vs Total Blood K

Note that Revici explicitly states on page 397 that he is not interested in the ratio of the potassium values (which is used in the calculation of the membrane potential as shown below). Unfortunately, he does not say why.
“It is not the ratio between these values which is of interest, but each value by itself.”

2 Our Interpretation

Our interpretation retains the use of serum and total blood potassium levels to indicate both relative potassium excess/deficiency and anaerobic/dysaerobic status. The difference lies in how the status is determined.

For relative excess/deficiency we use a calculation of total body potassium based on the ICF/ECF volumes and potassium concentrations. It will be shown that this depends almost entirely on the ICF (RBC) potassium concentration.

For anaerobic/dysaerobic status we use the potassium Nernst potential. This association is speculative, but is based on the importance of the cell membrane voltage physiologically and the fact that the primary determinant of the cell membrane voltage is the potassium Nernst potential across the cell membrane.

3 Total Body Potassium

Total body potassium is a straightforward calculation. \[ K_{tot} = 28L \cdot K_{ICF} + 14L \cdot K_{ECF} \] Here is a contour plot showing how total body potassium varies with the blood potassium measurements (the axes are similar to Figure 127 above).

We clearly see how total body potassium depends largely on RBC (intracellular) potassium.

Comparing this to figure 127 we see this interpretation of excess/deficiency is quite different. The quantitative value of \(K_{tot}\) may be useful for estimating how much potassium to supplement if deficient.

4 Potassium Nernst Potential

The Nernst potential for potassium (\(E_k\)) is calculated using the Nernst equation. \[ E_k = -V_t \cdot \ln \left( \frac{[K^+]_o}{[K^+]_i} \right) \] where \(V_t = \frac{kT}{q}\)
At body temperature (37C) this simplifies to \[ E_k = -27mV \cdot \ln \left( \frac{[K^+]_o}{[K^+]_i} \right) \]

Note that this is only an approximation of the actual resting membrane potential since it ignores the other ions (e.g. sodium and chloride). For perspective, in a neuron at rest \(E_k\) might be -86mV while the resting membrane potential was -65mV (Guyton and Hall, Textbook of Medial Physiology 9e, page 575). For further detail see the Goldman-Hodgkin-Katz Equation Calculator and notice that the resting membrane potential depends on the permeabilities and concentration gradient of each of the ions.

Using serum potassium as \([K^+]_o\) and RBC potassium as \([K^+]_i\) we create the following plot (again, the axes are similar to Figure 127 above).

We can see the correspondence between this and Revici’s Figure 127 Anaerobic and Dysaerobic quadrants. The general trend is the same from upper left (Dysaerobic, depolarized membrane) to lower right (Anaerobic, hyperpolarized membrane), but the classification of intermediate points is more precise.

We should be able to quantify the expected variation of \(E_k\) given that RBC potassium is (believed to be, over short time durations) a relatively stable measurement. Furthermore there is literature about the variation of serum K (e.g. circadian rhythm, gender, menstrual cycle).

Note that the usual expectation is that serum K and RBC K tend to vary together. It is important to notice that the most extreme values for \(E_k\) occur when this is not the case (i.e. one is high and the other low).

5 Case Study

A preliminary evaluation of this approach was done with a case study of a cancer patient over two years.

Here is a plot showing the time sequence of the patient’s potassium measurements. Contour lines and background color correspond to calculated resting membrane potential.

Plots like this can be used to track an individual’s current status and response to treatment.

A numerical table of the data:

test_date K RBC_K K_tot E_k
2011-10-12 4.60 77 2220.4 -76.07923
2012-05-01 4.20 79 2270.8 -79.22781
2012-07-27 4.50 87 2499.0 -79.96943
2012-10-15 4.40 84 2413.6 -79.62873
2012-12-05 4.50 100 2863.0 -83.72951
2013-01-25 5.00 88 2534.0 -77.43327
2013-03-20 4.35 87 2496.9 -80.88477
2013-06-01 4.10 80 2297.4 -80.21807
2013-09-12 4.30 89 2552.2 -81.81058
2013-11-25 4.10 86 2465.4 -82.17073
2014-12-26 4.50 87 2499.0 -79.96943

It is interesting to compare the ranges (for this patient) of the metrics and observe that \(K_{tot}\), \(K_{RBC}\), and \(K\) have a significantly greater percentage variation than \(E_k\). We believe this demonstrates that \(E_k\) is physiologically important and therefore under tight regulation. Small percentage changes in \(E_k\) are likely to be clinically significant. A key goal is to quantify the clinical significance of these small changes.
Range of \(E_k\): -83.7, -76.1 (variation of 10%)
Range of \(K_{tot}\): 2220, 2863 (variation of 25%)
Range of \(K_{RBC}\): 77, 100 (variation of 26%)
Range of \(K\): 4.1, 5 (variation of 20%)
The comparison with serum \(K\) is particularly relevant given that \(K\) is known to be under tight physiological regulation.

An interesting set of data points are the measurements from 2012-10-15, 2012-12-05, and 2013-01-25 showing dramatic changes which correspond to modifications in the treatment with Revici’s lipids. Of particular interest is the 6 mV change in \(E_k\) between 2012-12-05 and 2013-01-25.

6 Conclusions

Total body potassium (\(K_{tot}\)) and potassium Nernst potential (\(E_k\)) are physiologically meaningful quantitative values which may be useful as biomarkers.

The important question here is whether or not these biomarkers are clinically useful. The case study above is encouraging, but more data is needed to assess the utility of these biomarkers in clinical practice.

The next steps are:
1. Establish that these biomarkers can be measured effectively (e.g. lab accuracy and constincy, cost effectiveness).
2. Better understand the behavior of these biomarkers (e.g. reference ranges, circadian rhythm).
3. Link these biomarkers to physiological states and/or clinically relevant outcomes (e.g. hypertension, cancer).

Some preliminary work towards 2. appears in Revici Potassium Analysis. The focus is on leveraging the medical literature for serum potassium.

Revici’s work provides some validation of 1. and 3., but going further requires additional data. An expansion of the case study above to more subjects seems appropriate.

Subject data should include:

7 Additional Information

This document is a distillation of Revici Potassium Analysis. See that page for additional detail. Revici Table Index may also be helpful.

This analysis provides an example of working with a small serum and WBC K dataset.

See other documents in http://rpubs.com/rseiter for discussion of the Gubbio Study work on RBC and serum K.
Cohort profile: The Gubbio Population Study (M. Cirillo et al. 2013) gives a current (2013) overview of the study.
Red blood cell sodium and potassium concentration and blood pressure. The Gubbio Population Study (Trevisan et al. 1995) discusses RBC K in the context of hypertension.

File originally created: Monday, September 15, 2014
File knitted: Sun Jan 10 20:05:58 2016

Bibliography

Cirillo, M., O. Terradura-Vagnarelli, M. Mancini, A. Menotti, A. Zanchetti, and M. Laurenzi. 2013. “Cohort Profile: The Gubbio Population Study.” International Journal of Epidemiology 43 (3). Oxford University Press (OUP): 713–20. doi:10.1093/ije/dyt025.

Trevisan, Maurizio, Vittorio Krogh, Massimo Cirillo, Martino Laurenzi, Alan Dyer, and Jeremiah Stamler. 1995. “Red Blood Cell Sodium and Potassium Concentration and Blood Pressure.” Annals of Epidemiology 5 (1). Elsevier BV: 44–51. doi:10.1016/1047-2797(94)00040-z.