THE EFFECT of POTASSIUM CHLORIDE on HYPONATREMIA1 by JOHN H
Total Page:16
File Type:pdf, Size:1020Kb
THE EFFECT OF POTASSIUM CHLORIDE ON HYPONATREMIA1 By JOHN H. LARAGH2 (From the Department of Medicine, College of Physicians and Surgeons, Columbia University; and the Presbyterian Hospital in the City of New York) (Submitted for publication August 31, 1953; accepted February 10, 1954) In cardiac edema as well as in various other tration might favorably influence disturbances in states characterized by excessive retention of fluid sodium metabolism manifested by pathologic dis- there is observed frequently an abnormally low con- tribution of sodium and potassium within the body. centration of sodium in the serum and extracellular Two recent studies have served to emphasize this fluid (1, 2). Rigorous sodium restriction, mer- important relationship between sodium and potas- curial diuretics, and cation exchange resins may sium. In vivo, with potassium depletion there is exaggerate or produce this tendency to hypo- a movement of sodium ions into cells and an associ- tonicity. Sodium administration often enhances ated extracellular alkalosis. This intracellular so- accumulation of fluid, without increasing its to- dium can then be mobilized by potassium adminis- nicity, and hypertonic sodium chloride, though tration (11). In vitro it has been shown that po- effective at times, may be neither beneficial nor tassium transport is dependent upon energy of safe (3). aerobic oxidation. If the cell is injured or meta- Because of the shortcomings of these various bolically inhibited potassium fails to accumulate therapies and because hyponatremia per se may and is replaced by an influx of sodium and water play a role in the production of adverse symptoms, (12). it seemed desirable to search for another diuretic Accordingly, potassium chloride (KCl) was agent which might promote the loss of excessive given to an edematous, hyponatremic cardiac body water without aggravating disturbances in patient with the hope of effecting water diuresis. sodium metabolism. It was decided, therefore, to The procedure produced a reduced urine volume. observe the effects of potassium salt in hypona- Moreover, a striking increase in the serum sodium tremia, not only because of its possible action as a was observed at a time when there was no change diuretic agent but also because of its relation to in the external balance of sodium. This phe- sodium metabolism. nomenon seemed of sufficient interest to warrant Potassium salts have long been considered a further study. The observed and derived trans- useful diuretic agent. "Salt of niter" (potassium fers of sodium and potassium were, therefore, stud- nitrate) was recommended for the treatment of ied by the balance method following KCI ad- dropsy in 1679 by Thomas Willis (4). In 1920, ministration in six patients with and without Magnus-Levy (5) and Blum (6) showed that po- edema. tassium chloride could be administered safely by METHOD mouth in relatively large doses in various ede- Patient selection. Six patients, ranging in age from matous states and that it often produced a diuresis. 22 to 65 years, were studied on the medical or metabolic Similar results have been obtained by others (7-9). wards of the Presbyterian Hospital. The criteria for se- Potassium salts have produced diuresis without lection included hyponatremia (serum sodium less than excessive loss of sodium, acting apparently through 126 mEq. per liter) and the absence of advanced renal suppression of renal tubular secretion of hydrogen disease (blood urea nitrogen less than 30 mg. per cent, except for one value of 39 on one occasion in a single ions (10) and possibly as a result of osmotic ac- patient). One subject each with the following condi- tivity of electrolytes of tubular urine. tions was studied: rheumatic heart disease with edema, In addition to the renal effects of potassium Laennec's cirrhosis with ascites, tetralogy of Fallot salts, it was thought possible that their adminis- with edema, tuberculous meningitis, and unclassified col- lagen disease with accelerated hypertension and protein- I This work has been aided by a grant from the National uria. The latter two patients had no detectable edema. Heart Institute (USPHS). In addition, one patient was studied with edema due to 2 Research Fellow, New York Heart Association. cor pulmonale but without significant hyponatremia. 807 808 JOHN H. LARAGH SERUM 14or No EFFECT OF KCL ON SERUM Na 135 IN HYPONATREMIC SUBJECTS 1301 125I Legend 120 C --& Emq/L K 0-= 5 - 00 SERUM Ci 95 SERUM 130 No 90 12! 85 so SERUM 9* K 7 - SERUM 9 5 KCL 300 KCL 300 mEq 0 sEq . tO 20 30 0 10 20 DAYS .c. 3?(NDJ DAP eL Si5f congestive ,e/le,e ra ea I,g/II SEIRUM 135 SERUM 135, NW4w.. _ me' me 130 130 s2' 125 12 0 SERUM 120 IO SERUM 4L #$v*JLkt "0 Co Cl IQ5 I 105 100 N -~~~100 95 95 SERUM 9 SERUM 7 K 7 K s KCL 300 KIL 300 mLq O 10 20 30 40 0 10 20 DAYS DAYS or.P.f.aeC -V. Cvi,o,ele *-eeII. meV 06 FIGURE 1 Procedure. Four of the patients were transferred to and chloride. All stools were pooled for each period and the metabolism ward, where they were treated throughout analyzed similarly. the study with bed-rest and sedatives. No diuretic was Venous blood was drawn without the application of given, but digitalis therapy was maintained in constant a tourniquet when possible and was delivered under oil dosage. Weights were measured in the fasting state. for CO2 determinations. Blood specimens were usually The patients were maintained on constant amounts- of obtained daily in control periods and as often as six times distilled water and on a constant weighed, low sodium per day during periods of KCI administration. Blood- diet, the nature and quantity of which were selected by was allowed to clot for 45 minutes and then separated. the patient during a preliminary period. Serum was then analyzed for sodium, potassium, chlo- The balance studies were divided into periods of from ride, C02, urea, and total protein. In two patients, bal- three to six days during which the daily diet was kept ance studies were not done and serum electrolytes only constant by weight from identical food lots. Daily diets were observed. A 20 per cent solution of KCI was ad- were analyzed in duplicate periodically for sodium and ministered orally in amounts ranging from 54 to 295 potassium content and the composition of any vomitus or mEq. (4 to 22 grams) per day and given in divided uneaten residue was determined. Urine was collected dur- doses, three to four times a day. Serum potassiium levels ing consecutive 24 hour periods (8 a.m. to 8 a.m.), meas- were followed closely after each dose and serial electro- ured for volume, and analyzed daily for sodium, potassium, cardiograms were taken. THE EFFECT OF POTASSIUM CHLORIDE ON HYPONATREMIA 809 Analytical methods. Sodium and potassium were de- tween intake (calculated as the sum of diet and medi- termined with a barrier layer photo cell flame photometer cations) and output (the sum of losses in the urine and (lithium internal standard) having an accuracy of 1 stool). Because of the nature of the observed changes, per cent (13). Chloride was determined by the method no corrections were made for electrolytes lost by blood- of Schales and Schales (14) or by a modified Volhard letting, and the potassium balance was not corrected for titration. Serum carbon dioxide content was measured changes in nitrogen balance. No studies were carried out with the Van Slyke manometric apparatus. Serum urea in the summer months. nitrogen was determined by direct nesslerization accord- In considering possible changes in the distribution of ing to the method of Gentzkow (15). The hematocrit sodium and potassium between the extra- and intracel- was determined by the Wintrobe method, and serum pro- lular compartments, the intracellular balance was calcu- tein by a specific gravity gradient tube by Lowry's modi- lated as the difference between the measured external bal- fication of the micro-method of Linderstrom-Lang (16). ance and an estimate of the extracellular space of 20 per The exchangeable body sodium (24-hour) as well as the cent of body weight. This latter calculation is not es- 4-hour sodium space was determined employing Na' after sential to these studies but is used as a means of de- the method of Forbes and Perley (17). picting the type and direction of some of the changes. Analyses of diet and stool were made in triplicate. Since most of the patients were edematous, the assumed These specimens were homogenized and made up to a extracellular space is probably considerably less than the known volume with distilled water. Aliquots for sodium true value, and the postulated shifts of electrolyte are and potassium determinations were digested until clear therefore estimatediconservatively. For example, in two with equal volumes of concentrated nitric acid on an of the patients, J. C. and F. P. C., the four-hour Na' space electric hot plate, filtered, and made to volume with lithium was found to be 36 and 39 per cent of body weight, re- for analysis by flame photometer spectively. However, essentially the same conclusions Calculations. In the balance studies the total balance resulted from calculations based on larger initial extra- for sodium and potassium represents the difference be- cellular volumes. TABLE I Dosage schedules of KCI and observed effects on serum potassium Time for return to Duration Serum potassium control Total of the level Pt.* dose of dosage Observed (after last Date KCI period Controlt peak dose) mEq. days mEq./L. mEq./L. hrs. J. C. 187 2 4.5 9.5 120 3/7 J. C. 107 1.5 5.7 7.9 72 3/18 F.