intracellular cation exchanges in metabolic alkalosis
TRANSCRIPT
INTRACELLULAR CATION EXCHANGES IN METABOLICALKALOSIS
J. R. Elkinton, … , R. D. Squires, A. P. Crosley Jr.
J Clin Invest. 1951;30(4):369-380. https://doi.org/10.1172/JCI102453.
Research Article
Find the latest version:
https://jci.me/102453/pdf
INTRACELLULARCATION EXCHANGESIN METABOLICALKALOSIS 1
By J. R. ELKINTON,2 R. D. SQUIRES,3 AND A. P. CROSLEY, JR.4
(From the Department of Medicine and Hospital of The University of Pennsylvania,Philadelphia, Pa.)
(Submitted for publication October 30, 1950; accepted, February 5, 1951)
It is now well known that disturbances in theelectrolyte patterns of the intracellular fluid, aswell as in those of the extracellular fluid, may playa significant role in many disease states. Darrowand associates (1) have shown in experimentalanimals that one pattern of associated abnormali-ties in the two fluid phases involves intracellularpotassium deficit plus sodium excess and extra-cellular bicarbonate excess plus deficit of chloride.Since then evidence of this relationship has beenfound in a variety of clinical conditions includingcongenital alkalosis, vomiting, the postoperativestate, and syndromes of hyperadrenocortical func-tion such as Cushing's disease and ACTHadminis-tration. This evidence has been obtained by utili-zation of the balance technic during replacementtherapy (2-8), by study of the distribution ofradioactive sodium (9), and by tissue analyses ofskeletal muscle (10) and of erythrocytes (9, 11).These observations have suggested that both lowpotassium intake and hyperadrenocortical functionare implicated in the etiology of this disturbance.In addition to these two etiological factors a thirdhas been postulated, namely, impaired renal con-servation of the potassium ion (12).
The purpose of the present study is to examine,in a group of patients with metabolic alkalosis andhypokaliemia, 1) the direction and magnitude ofintracellular electrolyte abnormalities as revealedby the balance technic, 2) the relation of these ab-normalities to the extracellular hypochloremia andelevation of bicarbonate, and 3) the relative etio-
1 This work was done in the Chemical and Renal Sec-tions of the Department of Medicine and in the WilliamPepper Laboratory; the help of the laboratory staffs andof the clinical staff of the Hospital is gratefully acknowl-edged. Laboratory facilities were aided by grants fromthe National Heart Institute, U.S.P.H.S., and by theC. Mahlon Kline Fund of the Department of Medicine.
2Established Investigator of the American Heart As-sociation.
3Fellow of the Department of Medicine.4 Woodward Fellow.
logical roles of gastrointestinal fluid loss, low po-tassium intake, intrinsic renal dysfunction, andabnormal adrenocortical activity.
EXPERIMENTALMATERIAL ANDMETHODS5
Ten patients were studied by the balance technic, eightof whom were followed during fluid replacement therapyand correction of the alkalosis. The diagnoses of theseten patients and their pertinent clinical conditions arelisted in Table I.
Fluids and food of known composition were adminis-tered intravenously or orally. All excreta were collectedand analysed with the exception of formed stools whichusually contain insignificant amounts of electrolytes whenno salts have been administered by mouth. Sodium andpotassium were determined in serum, urine, and nitricacid digests of stools and foods by means of a Barclayinternal standard flame photometer according to themethods of Wallace, Holliday, Cushman and Elkinton(13). Chloride was determined in serum by the methodof Eisenman (14), in urine by the Volhard-Harveymethod (15); nitrogen was determined by micro- ormacro-Kjeldahl. Total CO2 content of serum and bloodwas measured by the method of Van Slyke and Stadie(16), and pH determinations were made with the glasselectrode or colorimetrically. Creatinine in urine was de-termined according to the method of Brod and Sirota(17) and in serum according to the method of Bonsnesand Taussky (18). Inulin determinations in serum andurine were made according to the method of Harrison(19); spaces and clearances being determined according
in-Out UVto the formulae, p and pW, respectively. The
method for the determination of urinary 17-ketosteroidswas modified from that of Holtorff and Koch (20).6
The derived data were calculated according to previousmethods (21). The fundamental assumption involvedis the equating of the extracellular fluid with the chloridespace; this assumption is discussed below. The changein chloride space is calculated from the change in con-centration of chloride in serum (corrected for serum wa-ter and the Donnan equilibrium), the balance of chloride,and an assumed initial or final volume for the chloride
5The laboratory assistance of Mrs. Helen Lightmanand Mrs. Claire Tissari is especially acknowledged.
6 The determinations of urinary 17-ketosteroids weredone in the Endocrine Laboratory under the direction ofDr. F. Curtis Dohan.
369
J. R. ELKINTON, R. D. SQUIRES, AND A. P. CROSLEY, JR.
(4
A4,9
t,.
Ei444
]'l 1t
.0 .
. i. ..i
pU
'0
0
0
0
V
B
.28
V
._
(4
0-
a..
I0
'0 %
o
go
0 Z_
co
0
0%
0
C4
0
..
4
to24)
0
0
o
z
0;.4
'0
*
'0te
ob
0
(4(40
.a
14
I 4
1I
U)0
(-4
(4a
cni
._
(4a
0
a
*!.4
(4
o(
04
ua
z z 4zt2 w t
0
Ioo o.I 1j
'0
00 ~ 0
0i.... 0
>4) 0 ..
o~~~~~~~~~CCo2
- -~~~e.-
N 2z 2 12 12 1oe~~~~~~~~~~~~~c
4.1 0 o 1 1
u
.5 ~ .~z
370
1.4
10
*0I> 84)0
4)
00
i.4)to 1-
4)94
zI
U'o
4.
0
(4)
4.I0
10
.0
r..
(.4
C:t
'0-
*;4
me
0a..2CD
I
I*'. .;
Qo c
0~
0- ?
zo
bi
9.
4)
4snc
4 _)
00
cd .
S co
(00 4)
.2.0 203 0
'0
Ui
(4
046.00
(43
._
O
(4.
=s
(4
I..
0)
10
(4L..
vC
0
eq
*8r.dJ9
. !
.8c
L.3
C4
(4.11
I I
INTRACELLULAR CATION -EXCHANGESIN METABOLIC ALKALOSIS
z* * *++ .1+.. +.. to
++ ++I+O I
oI ++
I I" I IIl I I
000'N4 -
I I
-H
,0++I .*w01 o t M0. rMOs -40o er- "> "W W))U* 0
+++ IT I++ ++ ++ ,1+ 1+ + ++ +++6 t N8o 0 ~008 'O'O UO 00- U) 0.) --
Z ++0O W- 000 + ++ '++ N++ ,0+7 +T j7+ +j ++ +++ + 7 ,
z
.
I++I-I.e
U) N++ I
-HV00 1 0- I I 0I -I 'N UtlCo I. I U) I I- -I I- *0-
U)+NN+ -++1+++1'++ +++ ++ +++ I++
l I I OI Ie>- I I'M'
NI W) ~H _01IVN N U)t o IU .
U) I4 WIn. )'itU I U) N0 U1w1)% 0 00Iq" in.1 * I'-I000 U2') 811) 0 W)'-I0'I.* N Nq U)U") - U)N N U)
wff~U* i0 *kI--- I- ° I
U I II I-I NoRC!%G U)0inG NU)j'.e i 4 ,00 mu)) CZ I . 00I- co
T1 o0leT -l I -I 400 U) V IU j)in InN I N in 4e_. N4 _U)
f~~~6j I~4- (4 Io S- 1" N40.-.£61 I>-00r 1 o lco jgjtoo I I M l O'b.U'O° It- e '0 U)O"I^ 1- It..+.. IN1 ° IN-
N Id,o 1. o o1 le4 1j 0Nff)|E""4|-1 ^+" 1 " 1 1 -| 1 * 1 * 14 -4
* U)U)C.
'CrI-1- 00C0
gv >8RvaW_ vx W)_4
1CUON"N
N _)U) -
I oU)coUin__Wt*r-I4__
- U)co0..On
-1(. .4
r _:
I * V)uV) .U no)U ) O O
Z CK2 eI 2 j00Got 10-4 ON co 4 ooo
]Off~OIo; " ) U>-c00 t- 00o0 'U O0 N"NU
I 00 N 1 8 VP Qr I ) 1 Ini1t -0 I .N co t-II--*--0 I*~ ~~'NOS'_-U1 ~ U)N1,E' NO1_0 1"01U1) '_ O.<~ ~ __NIs Io1o I^^I'-I I"-1°1 +Q en0o I )U) I 00 QN I UeI4-U ";4 U) a V;mI _U)U). t I-.WU) I NI. Qb '^0U) 1 U)0 1 ' U) e0I
bAtI- >I °I@j£O£1@ I -0 -~ON I)UOW I-U) 00U).
>b ~~~+
fi { { | | |U |M| | , £6|++ e g ++ i + + 2°:U
0 ~~~~0
0 I I 6:-wI_ o01eI^ _q Xo1ffi~~~~~~~~~~~~a + 8l 2 Sl|8|1vlA O- Tg1
+I +I I+1 I++ TI!-nj^^!^!> >!^!C!--"""!!+0r
61 I~~~~~a.. .IUo -I.C O10c -.U I.- .40 'ON N U) +M.N 'N INNN Nfm NNN N O~C4~I.0.- 0.CNN 00' .-N--N I~N t0 U)N~
[5 ~ Hi :3 :I
i 1-s1 1u(i :
1-
371
400a
'0
10
IT.
'-II0
0
.*
4)
._
O
'0
C:
co
._C3
04 4)
8 I
= _.
cO +
I-n
*4.)
0 IOCO°h
4!
-It
P4I
I.-
4)
:s
0
c
.oI-W4)
.0
cO
I.
I
++ I
i%O IV
A
J. R. ELKINTON, R. D. SQUIRES, AND A. P. CROSLEY, JR.
'0
--b go+4
in.6
Igo
-*r Wo -*00 It-e+b( --* 4 .*0+10 I I
*~~~~~~ ~ ~ ~~ ~-40- --qp vI .II-4
eni go.o e g) o --K P *- ot~~~ ~ ~ ~,W')ne *t ." ) .I C N
I+ + I++ ++ ++ +++ I+ ++ +1
qp 8 e44 W)t -4 MC4II
a-4 LI)gogc%b (4.
t4 eqm-I
lo Go'0go+ I
'a Go
I +
co6eq
gooceq go gogo eqIII
+4:go) go go0 # oV)o-
I I
go.,4
t1. eqeq...
I I
eqIeo
gogon0. o
eqg got'0 t.Wo .4
1 1
a eon
1
inogo++
s0goa.a'uin
(Pk en
II
-sgen++
'a do
1I+
.witqo0 )4Z d 4tim0v('z00t.c40 0,6'O o; CZ~ ('S0CUs 0.4g
0. o.aeqeqo 0 ntna t.. oen go go0% g40 eqgo
,, N gone en*00*'4en0 " "20~00~*0 I '* on'4@ .on
vwc_ v ""NI 0 - Nac @oeqe oe.m 0 o.e eqNK .. .go
_____ _ I4 '4-. - .4.4 - 4- " o
CD t.go000o00cc0 eqoo 1'- (k0 t m 0- G Coss
W)Gogo
IItqqe gogou1 eq googqI R C got goi It q
gogoeqeq0w=4@ W)m W -e o Goo e4m*vC0 0 M- u0g)~ Ot.t.
00~~ ii; lilt IIe'~~~~~ .e~~~jffW! q II I WIC' 'I.6.6o
>b ~~~~+A~~~~~~~~~~~~~~~~
0 mw0 ~~~~~~0 + .c 90 10+U' '~~~~~U+ ± m
tou+ .3 :-
)
10 I~~~~~~tnN 001 nf eqeo N 0 1 Io tnN
W)1In .. IO00 0M v1 "gogogo 0eqqe eqeN in
.4
'I 2!
I.0
I..
'0
(Us
a)10
S.-.
0
04
a)
0
(U03is
S..2la
'04C.)
(U
372
*
0
U
(.15s02
a)
0co
It0
"2
C4.
.0
a)0Q4
.0
Da
I.s
10
10-1
I
.4-4
.-I r..-4.-l 1
1 1
.i ,.; 4;-.4 .
INTRACELLULAR CATION EXCHANGESIN METABOLIC ALKALOSIS
space. The inulin space was determined in two patientsat the end of the period of study, but for reasons outlinedin the Discussicn were not used in the calculation ofchange in chloride space. From the values bbtained forthe volume of extracellular fluid and from the externalbalances of sodium and potassium, the exchanges of thesetwo ions between the extra- and intracellular phases ofthe total body fluids are calculated. The results of thislatter calculation are little affected by considerable varia-tion in the volume of extracellular fluid as originally as-
sumed or measured.
RESULTS
The results are presented in Tables I to IV, andin Figture 1. The clinical diagnoses of the patientsstudied are shown in Table I.
Intracellular cation exchanges. In all but one
(H. H.) of the ten patients potassium was re-
tained when administered (Table II). In thosepatients in whom it was possible to make the cal-culation, the retained potassium entered the intra-cellular phase in excess of nitrogen, in amountsup to 15.6 m.eq. per kilogram of body weight(Table III). During treatment, in seven out often patients, the balance of chloride was markedlypositive in relation to that of sodium (Table II).On the assumption that chloride exchanges were
extracellular, it is calculated that large amountsof sodium left the intracellular phase in thesepatients (Table III). In some of the patients(W. L., N. D., and M. F.) the increment of po-
tassium and the decrement of sodium were of simi-lar magnitude; in others, more sodium was lostfrom the cells than potassium was gained (TableIII). Only two patients were not shown to loseintracellular sodium during treatment, and in one
of these, A. J., the sodium exchanges were notmeasured.
Relation of intracellular cation exchanges toextracellular hypochloremia and bicarbonate ele-vation. In all the cases except J. C. the elevatedbicarbonate concentration in serum was associatedwith a lowered chloride concentration; and in allthe cases during treatment the latter rose as theformer fell (Table III). The high levels of bi-carbonate were not associated with any marked de-gree of hypernatremia; during treatment in allcases except M. F. the extracellular sodium con-
centration remained unchanged or rose. Duringcorrection of this hypochloremic alkalosis in eightof ten cases reported the associated intracellularcation exchanges were as follows. In four cases
TABLE IV
Relation of observed chloride exchanges to increments of chloride cakulated as necessary to reduce serum COCconcentration to 31 m.eq. per 1. (During early periods before response of alkalosis to therapy)
Serum Chloride
Patient PeriodCO, Cl Given Retained Excreted in Increment toC02ci Given Retained ~~~urine lower CO,*
days m.eq. per 1. m.eq. per 1. m.cq. meq. m.eq. m.eq.W. L. 0 45.0 79
2 40.1 88 247 200 45 66
H. H. 0 44.6 782 36.9 89 948 547 13 77
M. F. 0 40.0 882 45.0 82 82 16 5 97
N. D. 0 38.6 933 38.7 98 603 225 374 67
W. H. 0 52.8 623 40.1 95 165 96 72 12
A. J. 0 38.3 891 41.4 88 348 179 142 124
J. C. 0 31.0 1053 32.9 102 170 -146 307 27
* Calculated from the formula, E2X (ECO]2s2-31), where E2=extracellular volume at end of the period as derivedfrom an assumed value and the balance of chloride, in liters; ECOs2=2concentration of bicarbonate in serum at end ofperiod in m.eq. per 1.
373
J. R. ELKINTON, R. D. SQUIRES, AND A. P. CROSLEY, JR.
(W. L., N. D., W. H., and M. F.) to whompotas-sium was administered from the start of therapy,intracellular potassium increased and intracellularsodium decreased as the bicarbonate concentrationfell. In three others (A. K., H. H., J. C.) to whomNH4Cl was administered first, the bicarbonateconcentration fell to normal levels and sodium wasextruded from the intracellular phase in the ab-sence of any significant uptake of potassium. Intwo of these three cases (A. K. and J. C.) "excess"cellular potassium was subsequently taken up whenthe ion was administered; in the third, H. H., po-tassium was not retained when given. In theeighth case, F. G., the fall in bicarbonate concen-tration occurred in conjunction with a rise in intra-cellular sodium followed by a rise in intracellularpotassium. The dissociation in some of the cases
Serum ConC.: 45-Cot
A-A 35.OM per L
K t54.5'
o ° g3.5.e per"I.
Cumulative + *o00change in!
+4600'Cellulor K
in excess + 400'of N + too
of the uptake of intracellular potassium from thecorrection of the elevated bicarbonate concentrationand its frequent association with the extrusion ofintracellular sodium, as shown in three differentcases, is illustrated in Figure 1.
Etiology. The etiological factors responsible forthese electrolyte abnormalities are less readilyanalysed since in nine of the ten cases no observa-tions were made during the development of thealkalosis. A rough etiological classification ofthe cases has been attempted in Table I. (a) Low-K intake: Two cases (Group I) gave a history ofa low-K intake without any concomitant gastro-intestinal fluid loss. Both of these cases showeda maximal renal response for conserving potas-sium at the time the studies were initiated (see be-low), and neither case showed any evidence of in-
NH4CIKCI
--,0-0
KCI KCI
_A /K_
o.-
seq. oj-
Cellular
Na
- too'
- 400
-0o
- ool
r | f_S I_f __
R __l __,_
- ,^ .
| ,E __,
iWY_sRfli !O |
UF 9. A. K.
Der teu . . . . 10 O SSO O S
FIG. 1. RELATION OF CHANGESIN SERuMLEvzs OF CARBONDIOXIDE AND POTASSIUM TO CUMULA-TIVE CHANGESIN INTRACELLULAR CATIONS, DURING TREATMENTOF METABOLIc ALKALOSIS; THREETYPESOF RESPONSE
In Patient W. L. intracellular potassium increased and sodium decreased simultaneously with the returnof serum CO2 and K to normal levels. In Patient A. K. during administration of NH4C1 the serum CO,level returned to normal coincidently with a loss of cellular sodium and prior to an increase in intracel-lular potassium. In Patient H. H. the serum CO, level returned to normal in association with a loss ofintracellular sodium and without any significant subsequent uptake of cellular potassium.
374
4 KMP04- KCI - -t NH4Cl + * KCI -+
N..
\-,
&-.. -
INTRACELLULAR CATION EXCHANGESIN METABOLIC ALKALOSIS
creased adrenocortical activity. (b) Gastrointes-tinal fluid loss: Six cases (Group II) were knownto have extensive losses of gastrointestinal fluidcontaining potassium in addition to giving a historysuggestive of a low-K intake. Since, with the ex-ception of A. K., the gastrointestinal fluid lost in-cluded gastric fluid, probably greater amounts ofchloride were removed than of sodium. (c)Adrenocortical hyperfunction: In Group III arelisted two cases in which this factor apparentlyplayed a major etiological role. In M. F. the ad-ministration of large doses of ACTH, in conjunc-tion with a low intake of potassium, immediatelypreceded the onset of prostration associated witha severe alkalosis and hypokaliemia. This patienthad been shown during a previous course of ACTHto have an adrenocortical response as evidenced bya fall in circulating eosinophils, a rise in 17-keto-steroid excretion, and retention of sodium (22).In the other case, J. C., who had a typical adreno-genital syndrome due to carcinoma of the adrenalcortex, hyperfunction of the gland was the onlyfactor that could be adduced to account for the mildalkalosis and hypokaliemia found. In none of theother cases was there unequivocal evidence ofadrenocortical hyperfunction occurring as an etio-logical factor although thorough and systematicstudies of this aspect of the problem were notundertaken. In five of the other eight cases theexcretion rate of 17-ketosteroids was determinedand was found to be normal or low (Table I). Anegative nitrogen balance which might indicateadrenogenital hyperfunction was found in threeof these cases during treatment. Of these A. K.had an extensive loss of nitrogen from the bowel,N. D. was thyrotoxic, and H. H. was in the im-mediate postoperative period. It is therefore diffi-cult to use this observation as evidence of primaryadrenocortical hyperfunction. (d) Renal dys-function: kidney function was not systematicallystudied in this series of cases. Certain observationsin some of the cases, however, are worthy of note(Table I). Five cases did not have azotemia; two
of these, W. L. and F. G., were patients who ap-parently developed the alkalosis and potassium de-ficiency as the result of a low-K intake alone.During the first few days of study at the height ofthe alkalosis both of these patients excreted suchextremely small amounts of potassium that theU/P -ratio was less than 1.0. Conservation of po-
tassium, therefore, was maximal- at that time, de-spite the evidence from clearances of creatinineand inulin that glomerular filtration was impaired.
DISCUSSION
The observations made in this study confirmthe hypothesis that extensive disturbances of intra-cellular electrolyte exchanges usually occur instates of metabolic alkalosis. Intracellular deficitof potassium, as measured by uptake of the ionduring treatment, was certainly present in most ofthe cases. In Patient W. L. the deficit meas-ured in this way reached a maximum of 15.6 m.eq.per kilogram of body weight, and is of the samemagnitude as the largest deficits reported in in-fant diarrhea (23) or diabetic acidosis (24). Ifit is assumed that all of this potassium came fromskeletal muscle and that this tissue constitutes 40per cent of the body weight, the deficiency wouldbe equivalent to 39 m. eq. per kilogram of wetmuscle and would approximate 40 to 50 per centof the potassium in the intracellular phase. Sucha deficit is quite comparable to those produced ex-perimentally in rats by Darrow (1) and exceedsthose found in muscle biopsies taken from clinicalcases by Mudge and Vislocky (10).
The demonstration of an excess of sodium in theintracellular phase is less certain. As calculatedon the basis of the assumption that the chloride ex-changes were entirely extracellular, the intracellu-lar sodium exchange during treatment was nega-tive in all but one of the nine cases in which it wasmeasured. In four of the cases the size of thedecrement of cellular sodium was sufficiently closeto that of the increment of potassium to suggestthat the former cation was being substituted forby the latter. In the other four cases the loss ofcellular sodium so far exceeded the uptake of po-tassium as to obviate any such relationship. Thevalidity of the assumption, however, on which theseexchanges of intracellular sodium are calculated,must be questioned. Certainly in most of the casesduring treatment the external balance of chloridewas much greater in the positive direction than thatof sodium. These balances are observed, not de-rived, data, and would not be invalidated by anyunmeasured loss of salt in sweat in which chlorideand sodium are present in approximately equalconcentrations. The differential secretion of Clinto gastric fluid which was not lost from the body
375
J. R. ELKINTON, R. D. SQUIRES, AND A. P. CROSLEY, JR.
and therefore not included in the measured bal-ance, conceivably might be another source of er-ror. The magnitude of the discrepancies betweenCl and Na renders this explanation improbable; itis unlikely that, during the course of study in someof the patients, 5 or 6 1. of unreabsorbed gastricjuice accumulated undetected. The difference be-tween the external balances of these two ions wouldappear to require some other explanation. An al-ternative to the interpretation that these disparatebalances represented loss of sodium from the cellsis that chloride entered the intracellular phase.Physiological evidence for such a phenomenon isnot lacking. Conway's (25) theory calls for asmall amount of intracellular chloride to balancethe extracellular potassium, and Dean (26), Wilde(27), and Burch and co-workers (28) have pre-sented evidence that under certain circumstanceschloride may enter the cells of tissues other thangastric mucosa, testes, and ovaries where chlorideis a normal intracellular constituent. Evidence insupport of the entry of chloride into cells was foundin three of the cases reported here, W. H., W. L.,and A. K. In the infant, Patient W. H., duringtherapy the chloride space expanded to a volumeequivalent to approximately 50 per cent of thebody weight which is unexpectedly large for anon-edematous subject even in childhood. Asimilar finding has been described during treat-ment of vomiting infants by Danowski and hisassociates (4), who were unable to attribute it toerror in measurement of the chloride balance. InPatients W. L. and A. K. an inulin volume of dis-tribution was determined at the end of replace-ment therapy. When this volume of distributionwas taken as the final volume of extracellular fluidand, from it, and from the chloride balance, the ini-tial volume of extracellular fluid was calculated, thelatter was found to be less than 1 1. Since such avalue is absurd, the conclusion is that either themeasurement of the volume of distribution of inu-lin was in error, or that chloride entered cells.A definitive answer to this problem requires theaccurate measurement of the extracellular fluidvolume by some independent means, such as inulin,both before and after treatment, so that the intra-cellular transfer of chloride, if any, can be measureddirectly. Unfortunately, an opportunity to do thisin the present study was not available. The pos-sibility remains, therefore, that chloride entered
the intracellular phase during treatment in thesecases. But again because of the magnitude of thetransfers involved, it seems most unlikely that thisdid occur to such an extent as to preclude the ex-trusion of sodium.
Certain conclusions can be drawn in regard tothe relationship of these intracellular electrolyteabnormalities to the extracellular hypochloremiaand alkalosis. The production of intracellular po-tassium deficiency and of probable sodium excessby prolonged absence of potassium intake, appearedto result in hypochloremia and elevation of theextracellular bicarbonate concentration. This sug-gests that under conditions of biological equilib-rium or the steady state of adequate renal adjust-ment, as predicted by Darrow (1), these intra-cellular disturbances are associated with a dimin-ished tubular reabsorption of chloride despite thepresence of hypochloremia. Such a concept re-ceives further support from an analysis of therenal response to chloride administration in someof these cases, as presented in Table IV. In threeof the cases tabulated, the extracellular fluid vol-ume was expanding to such an extent that the ad-ministered chloride was insufficient to raise the ex-tracellular concentration even if the kidneys hadbeen able to conserve it, i.e., the kidneys excretedless chloride then the increment calculated as nec-essary to displace the bicarbonate to normal levels.In other words, in these cases, the kidneys had nochance to show conditioning by another factor andthe data cannot be used to imply that the circum-stances required a maintenance of hypochloremiaand alkalosis in the presence of chloride adminis-tration. In the other four cases of alkalosis, threeof whomhad hypochloremia (N. D., W. H., A. J.,and J. C.), during the initial days of study chloridewas excreted by the kidney in amounts sufficientto have displaced the high bicarbonate had thechloride been conserved. Thus the failure to ad-just the abnormal concentrations in the latter caseswas not due to unavailability of chloride when neu-tral salts, NaCl and KCI, were given. Pitts, Ayerand Schiess (29) have found that, during the acuteadministration of sodium bicarbonate, as the tubu-lar reabsorption of HCO8rises to a maximal rate,the reabsorption of chloride is correspondingly de-pressed; and as a result the total cation concentra-tion in the extracellular fluid is undisturbed. Thecases reported here are hardly comparable to these
376
INTRACELLULAR CATION EXCHANGESIN METABOLIC ALKALOSIS
experimental subjects of Pitts since ours representprolonged states of equilibrium maintained in thepresence of conditions other than acute loading ofexcess sodium bicarbonate. Furthermore, if theresults of the acute i.v. administration of a singledose of sodium bicarbonate to a normal subject areobserved over a period of many hours the initialrise in the rate of renal excretion of chloride is fol-lowed by a fall (30); as the elimination of the ex-cess bicarbonate continues the kidneys again con-serve chloride and the extracellular concentrationsreturn to normal. In the last four cases shown inTable IV, therefore, at a time when chloride wasavailable some other factors must have been con-ditioning the tubular reabsorption of this ion.
However, when chloride ion was administeredrapidly and in large quantities as an acidifying salt,NH,C1, chloride was retained in sufficient amountsto raise its extracellular concentration to normalas the extracellular bicarbonate concentration waslowered. Simultaneously, the extracellular volumeor chloride space expanded and sodium moved intoit from the intracellular space. This occurredwithout any restoration of potassium to the cells.There are at least two interpretations of thesetransfers. One is that an excess of intracellularsodium may have made this adjustment possible.Another is that the kidney was unable to excrete atleast part of the suddenly imposed load of extraanion with ammonia because of renal damage orbecause of the time lag inherent in the mechanismof ammonia formation. Under these circum-stances, the transfer of cellular sodium (the onlysodium available) was the only way to mitigate anotherwise abrupt change in acid-base equilibriumdue to displacement of bicarbonate as CO2 throughthe lungs. Such a sodium transfer requires someaccounting in regard to the intracellular anion pat-tern; this problem is discussed below. But which-ever of the above interpretations is accepted, of theresponse to NH,C1, it would appear that if thereis an obligatory relationship between the extracel-lular concentrations of chloride and bicarbonateand intracellular cation compositions, the relation-ship pertains more directly to intracellular sodiumthan to potassium.
According to the Darrow hypothesis, the directwithdrawal of chloride in excess of sodium by wayof gastric fluid, resulting in a metabolic alkalosis,should produce changes in intracellular cation com-
position even in the presence of an adequate intakeof potassium. Unfortunately our data are incom-plete on this point. Only in Patient T. W. werebalance studies obtained under such conditions.Intracellular potassium was diminished becauseof the loss in the gastric washings, but intracel-lular sodium was not increased. This may be ac-counted for by the presence of moderately severerenal insufficiency which certainly interfered withany renal readjustments that inight have beenmade. Prior to study, Patient H. H. had lost alarge amount of gastric fluid by lavage. Studiesduring his replacement therapy showed a verylarge decrement of cellular sodium as his bicar-bonate concentration fell, without much retentionof potassium. Thus under some circumstancesextracellular chloride deficit and alkalosis probablycan be produced without necessarily affectingeither the sodium or potassium content of the in-tracellular phase. This is in agreement with theexperience of Evans and co-workers who found inboth surgical patients and in dogs that alkalosiscould occur without cellular potassium depletionand could be corrected without the administrationof this ion (31-33).
Darrow has suggested that the primary factorin extracellular fluid which conditions the intra-cellular electrolyte pattern is the concentration ofhydrogen ion rather than that of bicarbonate orchloride. In support of this thesis he has citedhis experiments (34) in which rats on exposureto low atmospheric pressures developed sodiumexcess and a questionable potassium deficit in thecellular phase of skeletal muscle. Presumably analkalosis developed due to primary CO2 deficitwith a lowered CO2 content of extracellular fluid,although no such measurements were made. Wehave studied patients representing the converseof this condition, namely, those with respiratoryacidosis due to the primary retention of CO2 as aresult of severe pulmonary disease. This conditionleads to a secondary fall in chloride and rise in bi-carbonate concentration as a response to the risein level of carbonic acid. In a small series of suchpatients with marked elevations of serum bicar-bonate concentration, but with low blood pH, theserum concentration of potassium was found to beuniformly within the normal range (35). Thisfinding suggests, but does not prove, that no se-vere intracellular depletion of potassium coexisted
377
J. R. ELKINTON, R. D. SQUIRES, AND A. P. CROSLEY, JR.
with the high bicarbonate concentration found inrespiratory acidosis. These studies are not ade-quate controls to those reported in this paper sincethe uptake of potassium administered to such pa-tients was not tested. But they do suggest thatintracellular electrolyte composition is conditionednot by concentrations of extracellular bicarbonatebut by concentrations of hydrogen ion or carbonicacid. The latter is suggested by data of Wallaceand Hastings (36) who found by direct analysisof muscle that the bicarbonate content of intra-cellular fluid is very little affected by the bicarbonateof theextracellular phase (the cell is impermeableto HCO-), the calculated intracellular pH be-ing altered according to changes in CO2 pressurealone.
Extensive differential net transfers of cationsversus anions between the two phases of bodyfluids require consideration of their effect on theionic electrical neutrality and osmotic pressure ofthe respective phases. Knowledge of the ways inwhich electrical neutrality of intracellular fluid ismaintained under these circumstances awaits atleast the complete description of net exchangesacross the cell boundary of the cations, calcium andmagnesium, in addition to sodium and potassium,and of the anion, phosphate, as well as chloride.Some evidence on this score has been suppliedby Gardner, MacLachlan and Berman (37) whofound in the skeletal muscle of potassium deficientalkalotic rats, that the intracellular bicarbonatewas decreased and phosphate increased, and thatcalcium and magnesium, as well as sodium enteredthe cellular phase. However, in addition to ex-ternal transfers, metabolic processes and variationsin cellular pH must affect the degree of dissocia-tion of the organic phosphate and proteinate buf-fer system within the cell. Alterations in cellularosmolarity, so produced, apparently share in con-ditioning movements of water (38). Therefore,any complete accounting of the maintenance ofelectrical neutrality and of osmotic equilibrium isimpossible with the data at hand.
From the study of these cases it appears to theauthors that a variety of disturbances in intracel-lular cation pattern may exist in the presence ofmetabolic alkalosis, produced in a variety of ways.These disturbances may consist of (a) a decreasein potassium and an increase in sodium, (b) anincrease in sodium without change in potassium,
and (c) a decrease in potassium without an in-crease in sodium. The first of these, a, is mostcommonand is usually present when there has beenover-activity of the adrenal cortex or when the in-take of potassium has been very low for a pro-longed period of time. Under the latter circum-stance the kidneys continue to excrete potassiumfor a long enough period to produce a significantdeficiency before renal conservation takes place.Likewise, prolonged loss of gastric fluid, usuallyassociated with inadequate potassium intake, maybecome associated with a decrease in potassium andan increase in intracellular sodium. On the otherhand abrupt and drastic disturbance of extracellularfluid composition, such as may occur with exces-sive vomiting or postoperative gastric lavage, maybe associated with no intracellular changes or withother patterns of change such as b or c listed above.In such conditions are more likely to be found therenal changes described by Burnett, Burrows andCommons (12) leading to potassium wasting, andin such conditions is less likely to be found thesteady state of biological equilibrium (Darrow'sprerequisite to intracellular pattern, a). It wouldbe hazardous, therefore, to predict the precise in-tracellular cation disturbances in any case of meta-bolic alkalosis without due consideration of thephysiological and clinical circumstances of itsorigin.
SUMMARYAND CONCLUSIONS
By use of the balance technic the intracellularcation exchanges have been studied in ten casesof metabolic alkalosis during treatment.
In all but one of the cases potassium was retainedin the intracellular phase. In seven of the ten casesthe chloride balance was much more positive thanthat of sodium. On the assumption that chlorideis extracellular, large amounts of sodium were cal-culated to have left the intracellular phase. Insome patients the decrement of sodium was roughlyequivalent to, and in others greatly exceeded, theincrement of potassium.
The extracellular bicarbonate concentration fellto normal levels concurrently with the increase inpotassium and decrease in sodium in the intra-cellular phase, in four cases. In three others whenNH4C1 was given, the bicarbonate concentrationfell and sodium was extruded from the cells priorto any significant uptake of potassium by the latter.
378
INTRACELLULAR CATION EXCHANGESIN METABOLIC ALKALOSIS
In one of these cases no potassium was retained,and in one sodium entered cells as the alkalosis was
corrected.A variety of etiological factors appeared to be
involved in this group of cases as follows: lowpotassium intake alone, two; low potassium intakeplus gastrointestinal fluid loss, six; adrenocorticalhyperfunction (ACTH therapy, carcinoma), two.No evidence of abnormal renal wasting of potas-sium was found and in two cases an extreme de-gree of conservation of the ion with U/P ratiosof less than 1.0 was observed.
It is concluded from these data that althoughdecrease in potassium and increase in sodium is a
common intracellular cation disturbance in meta-bolic alkalosis, other patterns of disturbance may
occur, especially when the onset has been abruptand renal adjustment has not taken place. Wherethere is an equilibrium state between extracellularbicarbonate concentration and intracellular cationtransfers, the former is more directly linked tothose of sodium than to those of potassium.
REFERENCES
1. Darrow, D. C., Schwartz, R., lannucci, J. F., andCoville, F., The relation of serum bicarbonateconcentration to muscle composition. J. Clin. In-vest., 1948, 27, 198.
2. Darrow, D. C., Congenital alkalosis with diarrhea.J. Pediat., 1945, 26, 519.
3. Tarail, R., and Elkinton, J. R., Potassium deficiencyand the role of the kidney in its production. J.Clin. Invest., 1949, 28, 99.
4. Danowski, T. S., Austin, A. C., Gow, R. C., Mateer,F. M., Weigand, F. A., Peters, J. H., and Green-man, L., Electrolyte and nitrogen balance studiesin infants following cessation of vomiting. Pedi-atrics, 1950, 5, 57.
5. Lowe, C. U., Rourke, M., MacLachlan, E., and But-ler, A. M., Use of parenteral potassium therapy insurgical patients. Pediatrics, 1950, 6, 183.
6. Burnett, C. H., Burrows, B. A., Commons, R. R., andTowery, B. T., Studies of alkalosis. II. Electro-lyte abnormalities in alkalosis resulting from py-
loric obstruction. J. Clin. Invest., 1950, 29, 175.7. Kepler, E. J., Sprague, R. G., Clagett, 0. T., Power,
M. H., Mason, H. L., and Rogers, H. M., Adrenalcortical tumor associated with Cushing's syn-
drome. Report of a case with metabolic studiesand remarks on the pathogenesis of Cushing's syn-
drome. J. Clin. Endocrinol., 1948, 8, 499.8. Sprague, R. G., Power, M. H., Mason, H. L., Albert,
A., Mathieson, .D. R., Hench, P. S., Kendall, E. C.,Slocumb, C. H., and Polley, H. F., Observations
on the physiologic effects of cortisone and ACTHin man. Arch. Int. Med., 1950, 85, 199.
9. Eliel, L. P., Pearson, 0. H., and Rawson, R. W.,Postoperative potassium deficit and metabolic alka-losis. New England J. Med., 1950, 243, 471, 518.
10. Mudge, G. H., and Vislocky, K., Electrolyte changesin human striated muscle in acidosis and alkalosis.J. Clin. Invest., 1949, 28, 482.
11. Kennedy, T. J., Jr., Winkley, J. H., and Dunning, M.F., Gastric alkalosis with hypokalemia. Am. J.Med., 1949, 6, 790.
12. Burnett, C. H., Burrows, B. A., and Commons, R. R.,Studies of alkalosis. I. Renal function during andfollowing alkalosis resulting from pyloric ob-struction. J. Clin. Invest., 1950, 29, 169.
13. Wallace, W. M., Holliday, M., Cushman, M., andElkinton, J. R., The application of the internalstandard flame photometer to the analysis of bio-logical material. J. Lab. & Clin. Med., 1951, inpress.
14. Eisenman, A. J., A note on the Van Slyke methodfor the determination of chlorides in blood andtissue. J. Biol. Chem., 1929, 82, 411.
15. Harvey, S. C., The quantitative determination of thechlorids in the urine. Arch. Int. Med., 1910, 6, 12.
16. Van Slyke, D. D., and Stadie, W. C., The determina-tion of the gases of the blood. J. Biol. Chem., 1921,49, 1.
17. Brod, J., and Sirota, J. H., The renal clearance ofendogenous "creatinine" in man. J. Clin. Invest,1948, 27, 645.
18. Bonsnes, R. W., and Taussky, H. H., On the colori-metric determination of creatinine by the Jaffe re-action. J. Biol. Chem., 1945, 158, 581.
19. Harrison, H. E., A modification of the diphenylaminemethod for determination of inulin. Proc. Soc.Exper. Biol. & Med., 1942, 49, 111.
20. Holtorff, A. F., and Koch, F. C.. The colorimetricestimation of 17-ketosteroids and their applicationto urine extracts. J. Biol. Chem., 1940, 135, 377.
21. Elkinton, J. R., Winkler, A. W., and Danowski, T. S.,Transfers of cell sodium and potassium in experi-mental and clinical conditions. J. Clin. Invest.,1948, 27, 74.
22. Elkinton, J. R., Hunt, A. D., Jr., Godfrey, L., Mc-Crory, W. W., Rogerson, A. G., and Stokes, J.,Jr., Effects of pituitary adrenocorticotropic hor-mone (ACTH) therapy. J. A. M. A., 1949, 141,1273.
23. Darrow, D. C., The retention of electrolyte duringrecovery from severe dehydration due to diarrhea.J. Pediat., 1946, 28, 515.
24. Danowski, T. S., Peters, J. H., Rathbun, J. C., Quash-nock, J. M., and Greennman, L., Studies in diabeticacidosis and coma, with particular emphasis onthe retention of administered potassium. J. Clin.Invest., 1949, 28, 1.
25. Conway, E. J., Exchanges of K, Na and H ions be-tween the cell and its environment. Irish J. M. Sc.,1947, p. 593.
379
J. R. ELKINTON, R. D. SQUIRES, AND A. P. CROSLEY, JR.
26. Dean, R. B., Theories of electrolyte equilibrium inmuscle. Biol. Symposia., 1941, 3, 331.
27. Wilde, W. S., The chloride equilibrium in muscle. Am.J. Physiol., 1945, 143, 666.
28. Burch, G. E., Threefoot, S. A., and Ray, C. T., Ratesof turnover and biologic decay of chloride andchloride space in the dog determined with the long-life isotope, ClT. J. Lab. & Clin. Med., 1950, 35,331.
29. Pitts, R. F., Ayer, J. L., and Schiess, W. A., Therenal regulation of acid-base balance in man. III.The reabsorption and excretion of bicarbonate. J.Clin. Invest., 1949, 28, 35.
30. Singer, R. B., Clark, J. K., Crosley, A. P., Jr., andElkinton, J. R., Unpublished studies.
31. Evans, E. I., Potassium deficiency in surgical patients:its recognition and management. Ann. Surg., 1950,131, 945.
32. Van Slyke, K. K., and Evans, E. I., The paradox ofaciduria in the presence of alkalosis caused byhypochloremia. Ann. Surg., 1947, 126, 545.
33. Evans, E. I., Personal communication.34. Darrow, D. C., Changes in muscle composition in
alkalosis. J. Clin. Invest., 1946, 25, 324.35. Singer, R. B., and Elkinton, J. R., Unpublished data.36. Wallace, W. M., and Hastings, A. B., The distribu-
tion of the bicarbonate ion in mammalian muscle.J. Biol. Chem., 1942, 144, 637.
37. Gardner, L. I., MacLachlan, E. A., and Berman, H.,Carbon dioxide, inorganic phosphate and cationcontent of muscle in potassium deficiency. Feder-ation Proc., 1950, 9, 175.
38. Elkinton, J. R., Winkler, A. W., and Danowski, T. S.,Inactive cell base and the measurement of changesin cell water. Yale J. Biol. & Med., 1944, 17, 383.
380