Kidney International, Vol. 38 (1990), pp. 720—727

Disorders of distal acidification

NEIL A. KURTZMAN

Department of Internal Medicine, Texas Tech University Health Sciences Center, Lubbock, Texas, USA

The final site of titration of the urinary buffers is the collectingtubule. When this process fails urinary acid excretion is unableto match the 1 to 1.5 mEq/kg body weight of nonvolatile acidgenerated daily as a consequence of the high protein dietconsumed by most Westerners. The clinical expression of sucha defect is distal renal tubular acidosis (dRTA). Over the pastdecade numerous laboratory and clinical studies have clearlydemonstrated that dRTA is a group of syndromes which resultsfrom a variety of defects in distal urinary acidification.

If 100 mEq of nonvolatile acid are generated a day, the kidneymust reclaim all of the filtered bicarbonate and excrete 100 mEqof acid to preserve homeostasis. Since the urine pH cannot fallbelow about 4.5, the acid must be excreted in buffered form astitratable acid or as ammonium. Though the nephron canreabsorb bicarbonate along its entire length, bicarbonate recla-mation is mainly a proximal event. Similarly, the tubule titratesthe urinary buffers at all sites. The collecting tubule, however,is the location of the greatest lowering of urinary pH and bearsthe major responsibility for acid excretion.

In this paper I will discuss the mechanisms that might beresponsible for the clinical syndromes of renal tubular acidosisand present a classification of these disorders. Many of thesemechanisms can only tentatively be offered as a model since ourknowledge is not far enough advanced to be certain about them.The model will doubtless require modification as more informa-tion is obtained.

Mechanisms of distal acidification

Acid is secreted in both the cortical and medullary collectingtubules. The intercalated cells of these segments clearly havebeen implicated in this process [1]. These cells contain carbonicanhydrase as well as an H-ATPase which, depending on itslocation, mediates proton or bicarbonate secretion, the formerwhen it's on the apical membrane, the latter when it's on thebasolateral membrane. There is also evidence, however, thatthe principal cell of the cortical collecting tubule (CCT) partic-ipates in acidification [2], at least under some circumstances.

Oubain stimulates CCT acidification in tubules that have hadcarbonic anhydrase inhibited, as does basolateral amiloridewhen applied to CCT from starved rabbits [2]. The oubain effectis blocked by luminal amiloride. These data indicate that thesodium transporting cells, the principal cells, acidify the urine.This interpretation is based on the following assumptions.Carbonic anhydrase inhibition abolishes acidification in the

© 1990 by the International Society of Nephrology

intercalated cells, thus the stimulation of acidification seen iithis circumstance following oubain must be due to an effect otthe principal cell. In addition luminal amiloride should onl3affect the cell with a sodium channel. By raising cell sodiumoubain would decrease basolateral sodium-hydrogen exchangewhich then would stimulate apical proton secretion. The nexassumption is that basolateral amiloride is blocking a basolateral sodium-hydrogen exchanger in a sodium transporting cellAs was true for oubain, this would lower cell pH and stimulatapical proton secretion.

In addition to the H-ATPase, data are accumulating thasuggest that another proton pump also modulates collectinltubule proton secretion, the H-K-ATPase [3, 4]. This transporter appears identical to the gastric acid pump and is inhibitable by omeprazole and vanadate. It is stimulated bpotassium depletion. Figures 1 to 3 depict the model of hydro•gen secretion postulated in the collecting tubule. Before dis.cussing it, however, I must caution that much of it is theoreticaland requires experimental verification before it can be accepted.

The CCT, according to the model, contains principal celbwhich actively transport sodium and which contain the H-KSATPase at the luminal membrane, while a sodium-hydrogenexchanger is at the serosal surface (Fig. 1). Potassium depletionwould enhance acid secretion by this transporter. Thus thepostulated role of the H-K-ATPase in this cell is to regulatepotassium absorption and cell PH; it would also contributesomewhat to overall acid-base regulation. The source of theprotons extruded at both plasma membranes is likely metabolicacid which does not require carbonic anhydrase for its genera-tion. There is also a chloride-bicarbonate exchanger on thebasolateral membrane (not shown on Fig. 1) which will extrudebicarbonate only when luminal proton secretion exceeds baso-lateral proton transport [5].

The CCT also has intercalated cells, though in much lowernumber, which contain the H-ATPase and, I believe, theH-K-ATPase as well [1]. Some of these intercalated cells havethe H-ATPase at the apical membrane (alpha cells) (Fig. 2),while others have the pump at the basolateral surface (betacells) (Fig. 3). The beta cell also appears to have a sodium-hydrogen exchanger at its basolateral membrane (not shown)[5]. I think it likely the H-K-ATPase is present only in the alphacells. The medullary collecting tubule (MCT) contains onlyprincipal cells and alpha intercalated cells [1].

Aldosterone stimulates the sodium pump (whether directly orindirectly is irrelevant to the model), increases passive potas-sium secretion, and increases H-ATPase activity. It does not

720

Kurtzman: Disorders of distal acidfficarion 721

H

H

Na *

stimulate the H-K-ATPase. Thus aldosterone excess is associ-ated with potassium wastage and excess acid secretion whilealdosterone deficiency is associated with potassium retentionand acid retention.

This model explains a variety of clinical and laboratoryobservations which have been difficult to integrate. For exam-ple, the combination of hypokalemia and aldosterone excess isassociated with increased acid excretion and metabolic alkalo-sis, while potassium depletion alone is usually accompanied bymetabolic acidosis. Steroid excess would stimulate CCT so-dium absorption which would increase lumen negative voltageand thus enhance electrogenic (H-ATPase) driven acidification.Steroid excess would also stimulate the H-ATPase directly inboth CCT and MCT and further increase acid excretion. Potas-sium depletion would stimulate the H-K-ATPase as well, fur-ther enhancing acid excretion.

By contrast, potassium depletion alone would decrease aldo-sterone release thus inhibiting the H-ATPase both directly andindirectly via voltage. The stimulatory effect of potassiumdepletion on H-K-ATPase would mitigate the loss of H-ATPaseactivity, explaining the variable acid-base status which charac-terizes patients and animals with pure potassium depletion.

Sodium sulfate infusion and furosemide administration bothstimulate distal acidification [6, 7]. This has been attributed to

HC03 increased lumen negative voltage in the CCT. The furosemideeffect is blocked by amiloride, which further supports thevoltage hypothesis [7]. An additional explanation is possible,however. Furosemide (and sulfate as well) might stimulateacidification by increasing potassium delivery which wouldenhance electrically neutral H-K exchange. Amiloride wouldblock this effect by lowering cell sodium concentration whichwould enhance basolateral sodium-hydrogen exchange, whichin turn would reduce luminal proton secretion (Fig. 1).

This mechanism does not exclude a sodium dependent effectof furosemide or sulfate on acidification. Under this scenarioincreasing distal sodium delivery with an poorly reabsorbableanion, that is, sulfate or chloride in the presence of furosemide,would increase CCT lumen negative voltage and enhance bothproton and potassium secretion.

There are at least five mechanisms that could fail and causedefective distal acidification, each of which will be discussed indetail below (Table 1). Either of the proton pumps might bedamaged. A defect in the H-K-ATPase should cause hypokale-mic dRTA. This might be the mechanism of classical dRTA. Ifthe H-ATPase were deranged a normokalemic form of RTAshould eventuate. Impaired CCT sodium transport would im-pair voltage and cause hyperkalemic acidosis. Similarly, aldo-sterone deficiency or resistance would also cause hyperkalemicacidosis. Finally, an inability to maintain a steep pH gradientacross the distal nephron would also cause metabolic acidosis.

Tests of urinary acidification

Bicarbonate reabsorptionSince most of the filtered bicarbonate is reabsorbed in the

proximal tubule, measurement of bicarbonate reabsorption is atest of proximal acidification [8]. Expression of bicarbonate

Fig. 1. Theoretical model of principal cell acidification. There is also achloride-bicarbonate exchanger and a potassium channel on the baso-lateral membrane that have been omitted for simplicity's sake.

Fig. 2. Model of the alpha (proton secreting) cell of the collectingtubule. CA = carbonic anhydrase

Lumen ..-.. Blood u nLumen Blood

HCO3Na -

Cl .

Fig. 3 . i/vt/el of the beta (bicarbonate writ- ti,i tell.

bicarbonate-secretingintercalated cell

+ HCO3 -

ATP

ADP-Pi CI-

F-I

AlP

AOP - Fl

proton-secretingintercalated cell

Lu men Blood

722 Kurtzman: Disorders of distal acidification

Table 1. Potential causes of dRTA

Defect Syndrome

H-KATPase activity Hypokalemic dRTAH-ATPase activity Normokalemic dRTACCT Na transport Voltage-dependent

(hyperkalemic) dRTAAldosterone deficiency or Hyperkalemic metabolic acidosis

resistance with low urine pHCapacity to maintain steep pH Hyperchloremic metabolicgradients acidosis with normal UpCO2a

a The hypokalemia typical of amphotericin induced dRTA is due toincreased cell permeability to potassium rather than decreased potas-sium transport capacity.

reabsorption per liter of glomerular filtrate facilitates interpa-tient comparison. Measuring bicarbonate reabsorption over awide range of plasma bicarbonate concentrations provides themaximum amount of information about proximal bicarbonatereclamation. Because it requires infusing bicarbonate overseveral hours, the test is seldom used clinically. Much easier,and therefore more widely used, is fractional bicarbonateexcretion. Maximal bicarbonate reabsorption is normal in pa-tients and animals with defective distal acidification. Sincesmall amounts of bicarbonate are reabsorbed in the collectingtubule, fractional bicarbonate excretion is increased in dRTA.It usually is 5 to 10%. An excretion of more than 15% indicatesa proximal defect.

Urine pCO2When the urine is rich in bicarbonate the urine pCO2 reaches

a value about 40 mm Hg or more greater than that of the blood.This rise is predominately the consequence of collecting tubuleproton secretion [9—111. Since luminal carbonic anhydrase isnot present in this segment the titration of bicarbonate gener-ates carbonic acid which dehydrates to CO2 and water slowly.The CO2 thus produced is formed in the terminal nephron andin the lower urinary tract where the surface to volume relation-ship is unfavorable for back diffusion. Impairment of distalproton secretion, irrespective of the mechanism, will impair thekidney's ability to increase urine pCO2 during bicarbonatediuresis. Thus this test is very sensitive and equally nonspecific.To yield interpretable results sufficient bicarbonate should beadministered to raise urinary bicarbonate concentration to atleast 100 mEq/liter and the urine pH should exceed 7.8 for atleast 30 minutes prior to the collection of the test sample.

Urine pHThe normal kidney promptly excretes an acid urine in re-

sponse to acidemia [121. When the blood pH falls below 7.35 theurine pH almost invariably is less than 5.5; the only exceptionsare impaired distal delivery of sodium and defective distalacidification. The former can easily be detected by measuringurine sodium concentration.

Sodium sulfateFailure to lower urine pH during acidemia because of dimin-

ished distal sodium delivery can be eliminated by infusingsodium sulfate [121. This procedure results in a prompt loweringof urine pH and in an increase in both potassium and ammo-

nium excretion. The mechanisms that may mediate this re-sponse and that to furosemide have already been discussed. Astypically performed, 500 ml of a 4% solution are infused overabout an hour. Urine collection should continue for two to threehours since some patients have a delayed response.

Furosemide administration

Although the sodium sulfate test provides much useful infor-mation, it is cumbersome and difficult to perform. The recentlydeveloped furosemide test gives the same information muchmore conveniently [7]. A dose of 40 to 80 mg of furosemide isgiven orally. Urine is collected 120 to 180 minutes later whenthe maximum acidification response occurs.

Urinary ammonium and the urine anion gap

A normal renal response to acidemia includes an increase inurinary ammonium excretion. If ammonium excretion goes upunder this circumstance renal acidification is intact regardlessof whether urine pH falls or not. Conversely, urinary acidifica-tion is impaired if it does not rise irrespective of how low theurine pH falls. Urinary ammonium is not usually measured inpatients. The amount of ammonium in urine can be estimated,however, by determining the urine anion gap, define as (Na +K) — Cl [13]. A urine rich in ammonium will have a negativeanion gap, ie there will be considerably more chloride presentthat sodium plus potassium; the missing cation is ammonium.

Renal tubular acidosis syndromes

Classical dRTA

As originally described, dRTA was a syndrome of hypokale-mia, hyperchloremic metabolic acidosis, inability to lower urinepH below 5.5, nephrocalcinosis and nephrolithiasis, and osteo-malacia or renal rickets [14—16]. Until about a decade ago thedisorder was thought to have a single pathogenesis: an inabilityto generate and maintain a steep pH gradient in the distalnephron [17]. In other words, the distal nephron was thought tosecrete protons normally, but because the tubule was "leaky"the acid back-diffused across the terminal nephron and wasreturned to the blood.

A series of animal experiments cast doubt on this hypothesis.As previously mentioned the urine pCO2 is a very sensitivemeans of assessing distal hydrogen ion secretion. If a gradientdefect were responsible for classical dRTA then the urine pCO2should rise normally in afflicted subjects, since the urine pH ishigh during bicarbonate diuresis and prevents a lumen to bloodacid gradient. This was confirmed by experiments in amphoter-icin treated rats which showed the capacity to generate CO2 tobe intact [18]. Earlier experiments in turtle bladder, a tightepithelium and an analogue of the collecting tubule, had shownthat amphotericin interferes with acidification solely by increas-ing acid back-diffusion [19]. Thus amphotericin is a model ofgradient dRTA. When patients with dRTA were studied theywere shown to have an impairment of urinary CO, generationduring bicarbonate loading [20]. Thus, their defect appeared tobe one of impaired proton secretion rather than increased acidback-diffusion.

At present the only form of dRTA thought to be secondary toa gradient defect is that induced by amphotericin. I mustemphasize, however, that almost all the pertinent amphotericin

Kurtzman: Disorders of distal acidification 723

Table 2. Features of voltage-dependent dRTA

1. Salt wastage—detectable only with marked salt restriction2. Hyperkalemic hyperchloremic metabolic acidosis3. Normal or increased plasma aldosterone4. Low urine pCO2 with bicarbonate loading5. Urine pH >5.56. Abnormal response to sulfate or furosemide administration

data come from animal studies. For this issue to be completelyresolved urine pCO2 data in patients with amphotericin neph-rotoxicity are needed.

The best explanation currently available for the hypokalemicform of dRTA is to postulate inhibition of the H-K-ATPase.Such a mechanism has the virtue of explaining all the features ofclassical dRTA, but rests on an inadequate data base. If thisenzyme were a major regulator of both potassium and acid-basehomeostasis then its failure would cause hypokalemic hyper-chloremic metabolic acidosis with all the feature associatedwith classical dRTA.

Examination of Figure 1 shows that failure of the pump woulddivert metabolic acid produced in the principal cell to thecirculation, and cause potassium wastage. The failure of theH-K-ATPase in alpha intercalated cell as well would cancel anyadaptive response by the H-ATPase and thus ensure the devel-opment of systemic acidosis.

Irrespective of the etiology of this syndrome, relentless acidretention rapidly exhausts extracellular buffers. Blood pH canthen only be defended by tapping another source of buffer; theonly pool big enough to satisfy this need is the skeleton.Accordingly, basic calcium salts are continually leached frombone. The clinical sequelae are hypercalciuria, nephrocalcino-sis, and nephrolithiasis as well as osteomalacia or renal ricketsdepending on the age of the patient.

Voltage-dependent RTAObstruction of a ureter for as little as 24 hours results in an

acidification defect that is apparent immediately following reliefof the obstruction [211. The lesion is also characterized byimpaired potassium excretion and reduced mineralocorticoid-stimulated sodium reabsorption. For these reasons the collect-ing duct was implicated as the major site of involvement.Because administration of the potassium sparing diuretic amil-oride caused an identical defect, amiloride was though to impairthe same mechanism as urinary tract obstruction [22].

Studies with amiloride in the turtle bladder revealed that thedrug interfered with acidification only indirectly [23]. When thetissue was voltage clamped amiloride had no effect on protonsecretion. In the open circuited state, however, it reducedacidification in tandem with its reduction of the transepithelialpotential difference. This suggested that there was a majorvoltage dependent component of acidification. The CCTseemed to be the logical site of this mechanism. Experiments inrabbit collecting tubule were consistent with this view [24].Amiloride and oubain inhibited acidification in CCT, but not inMCT where active sodium transport was said not to occur.

A syndrome of voltage dependent dRTA was postulated toexist [25]. Its features included hyperkalemic hyperchioremicmetabolic acidosis, an inability to lower urine pH, and a normalor elevated plasma aldosterone concentration (Table 2). Shortly

thereafter patients with urinary tract obstruction who fulfilledthese criteria were reported [261. Patients with hemoglobin Swere also shown to have an identical impairment of both acidand potassium excretion [27].

More recent studies from our laboratory suggest that theeffects of urinary tract obstruction on acidification are morecomplicated than we initially thought. Collecting tubules takenfrom the post-obstructed rabbit kidney showed decreased acid-ification only in the MCT [28]. In fact, acidification wasincreased in CCT after 24 hours of obstruction. Obviously, thisfinding is not consistent with a voltage mechanism.

Study of H-ATPase activity post-unilateral ureteral obstruc-tion showed a decrease in enzyme activity in both CCT andMCT, but the decline was much more profound in the latter[29]. Na-K-ATPase activity was depressed along the entirenephron. This study suggests that, in addition to a voltagemechanism, obstruction directly affects the (or a) proton pumpin the MCT. Finally, preliminary data from our lab indicatesthat amiloride administration inhibits H-ATPase activity in bothCCT and MCT [30]. This suggests that amiloride has a mecha-nism of action different from that postulated above or thatsodium dependent acidification occurs in the MCT as well asthe CCT.

Lithium administration to animals in toxic amounts causesdRTA [31]. This disorder is not associated with hyperkalemia,though the defect in acidification induced by lithium in the turtlebladder is identical to that caused by amiloride [32]. Weexplained this lack of hyperkalemia by postulating that lithiuminterfered with potassium reabsorption proximal to the CCT.The finding of a voltage-induced impairment of proton secretionshould be associated with a decrease in H-ATPase activity inthe CCT, which is what we recently found [301. As withamiloride, we also found a decrease in MCT as well. Thisdecrease in H-ATPase activity in both CCT and MCT isconsistent with the decrease in urinary pCO2 observed inpatients receiving therapeutic doses of lithium [33]. It alsosuggests that lithium interferes with acidification by doing morethan just reducing CCT transepithelial voltage.

What's lacking in all of these studies is the role of theH-K-ATPase in mediating both normal and deranged acidifica-tion. For example, urinary tract obstruction and amiloridemight decrease H-ATPase activity while stimulating the H-K-ATPase. Depending on the degree of stimulation and inhibitionof the two proton pumps overall acidification and potassiumexcretion might both be reduced.

Lithium, on the other hand, might reduce H-ATPase activitywithout stimulating the H-K pump. This would explain theabsence of hyperkalemia in lithium treated animals and itsgreater (than amiloride) propensity to causes systemic acidosis.Studies of the H-K-ATPase currently underway will likelyresolve these questions.

Normokalemic dRTA

Some patients with moderately advanced renal insufficiencyhave dRTA with neither potassium wastage nor retention [20].This syndrome also is seen in patients with chronic transplantrejection [34]. I think its cause is impaired H-ATPase activity inthe collecting tubule. If this enzyme failed, but the H-K-ATPaseand the Na-K-ATPase were both intact then normokalemicdRTA should ensue.

724 Kurtzman: Disorders of distal acidification

This syndrome is almost always acquired, and is commonlyseen in patients with interstitial nephropathy who usually havesubstantial reductions in glomerular filtration rate. Becauserenal insufficiency markedly reduces urinary calcium excretion,these patients do not have nephrolithiasis or nephrocalcinosis.Both the hypokalemic and normokalemic forms of dRTA caneasily be treated with the daily administration of 50 to 100 mEqof bicarbonate or citrate.

Rate dependent-RTAThere is a group of patients with reduced renal function

(usually due to interstitial nephropathy) who have a subtledefect in distal acidification [35]. They cannot raise the urinepCO2 in response to a bicarbonate load; in all other respectstheir ability to acidify appears normal. They typically do nothave acidosis. We termed this disorder rate-dependent dRTAand proposed that the definition of incomplete dRTA be broad-ened to include patients without overt acidosis who cannot raisethe urine pCO2, but who can lower urine pH. Our reasoning wasthat the urine pCO2 reflects the rate of distal acidification andthat a moderate decrease in that rate might be reflected only bya decreased urine pCO,. Only a handful of patients withrate-dependent dRTA have been described. Whether the defectis of more than marginal clinical importance is uncertain. It islikely that the syndrome represents an early stage of one of theother forms of RTA that may ultimately progress to an overtstate.

Aldosterone deficiency and resistance

Although originally described more than 30 years ago [36],the syndrome of isolated aldosterone deficiency has been rec-ognized with regularity only during the past decade and a half[37—40]. The hallmarks of the disorder are hyperkalemia andhyperchioremic metabolic acidosis. Because aldosterone exertsits effects on electrolyte transport on the collecting tubule thedisorder had been grouped with the renal tubular acidoses; it iscommonly called type IV dRTA. Since type I and type II areinfrequently used to describe distal and proximal renal tubularacidosis, since there is no type III, and since aldosteroneresistance and voltage dependent dRTA have similar featuresand are both often called type IV RTA, the term lacks precisionand should be discarded. Aldosterone deficiency is preferableas it identifies the primary defect and suggests the appropriatetherapy.

As noted above, aldosterone directly stimulates collectingtubule acidification by increasing proton secretion and indi-rectly by enhancing sodium absorption. Thus its absence resultsin acid retention; however, the ability to generate and maintainpH gradients across the distal nephron is not reduced in thiscase. The ability to lower urine pH in the face of aldosteronedeficiency is generally attributed to decreased urine ammoniageneration which is characteristic of the syndrome [41, 42]. Itdoesn't take much to lower urine pH when urinary bufferconcentration is low.

Aldosterone deficiency should slow the overall rate of distalproton secretion. To the extent that the urinary pCO2 reflectsthis parameter the data are problematic. We have been unableto show the aldosterone deficiency lowers the urine pCO2 [41,43]. DuBose and Caflisch [18], however, have shown a de-crease. Again, the missing piece of the puzzle may be the effect

of aldosterone on the H-K-ATPase. In the absence of aldoste-rone H-ATPase activity should markedly fall. That of theH-K-ATPase should be unaffected. Persistent H-K-ATPaseactivity may play a role in the amiloride inhibitable fraction ofurinary acidification seen in aldosterone deficient subjects [43}.

When amiloride is given to adrenalectomized animals theurinary pCO2 falls dramatically [43]. This demonstrates thatthere is a major component of distal acidification that isaldosterone independent and sodium dependent. This aldoste-rone independent component of acidification appears to medi-ate the response to saline loading in aldosterone deficientanimals. This maneuver completely corrects the acidosis ofmineralocorticoid deficiency and underlies the favorable re-sponse of patients with the disorder to diuretic therapy and saltadministration [41]. When distal delivery is high, aldosteronedeficiency is virtually without manifestation. When it is reducedhyperkalemia and metabolic acidosis are obvious.

While the story of aldosterone dependent and independentacidification is still incomplete, it is absolutely clear that pa-tients with pure aldosterone deficiency can lower the urine pHnormally during acidemia [26]. This capacity is extremely usefulin distinguishing this syndrome from voltage dependent renaltubular acidosis.

Aldosterone resistanceFor every endocrine deficiency disease there is a resistance

state [40]. The features of the resistance state are the same asthose of the deficiency syndrome, except that the hormone inquestion (in this case aldosterone) is present in normal orincreased concentrations. Thus the blood levels in aldosteroneresistance and voltage-dependent dRTA are the same. Thedistinguishing characteristic is that patients with aldosteroneresistance can lower their urine pH normally, while those withvoltage-dependent dRTA cannot.

A form of aldosterone resistance is seen in children that isassociated with profound salt wastage far in excess of thedeficiency or resistance syndromes of adults [44—46]. Thedifference may attributable, at least in part, to the normalglomerular filtration rate present in these children.

Another form of adult aldosterone resistance occurs in adultsthat presents with salt retention rather than wastage; its patho-genesis has been attributed to enhanced chloride permeabilityin the CCT [47, 48]. If chloride is reabsorbed across the CCT inincreased amounts ( a so-called chloride shunt), the tubulecannot sustain the negative charge that usually prevails. Theloss of this charge, in turn, secondarily reduces both proton andpotassium secretion. Excess salt reabsorption also generatesvolume-dependent hypertension. Patients with this disorderrespond normally to sulfate or furosemide administration, un-like those with voltage-dependent dRTA.

Evaluation

The approach to a patient with hyperchloremic metabolicacidosis begins with the examination of the urine. If the acidosisis the result of an extrarenal disorder, such as diarrhea, theurine will be rich in ammonium. This can easily be disclosed bymeasuring the urinary electrolytes. There will be considerablymore chloride than sodium plus potassium, usually >50 mEq/liter. In other words, the anion gap will be minus 50 or more.Patients with dRTA or related syndromes typically have more

Kurtzman: Disorders of distal acidification 725

Clinical approach to the classificationof renal tubular acidosis

Hyperchloremicacidosis with

normal or lowplasma potassium

Urine pHduring acidosis

I I I I

Exclude proximal renaltubular acidosis,

measure ammonium and Diagnose distaltitratable acidity, renal tubular

measure urine PCO2 acidosis ifduring sodium bicarbonate

bicarbonate infusion excretion is(i.e., look for a "rate not highdependent" defect(

Fig. 4. Class(fication of RTA in patients with normal or low potassiumconcentration. Adapted from reference 49.

Hyperchloremicacidosis withhyperkalemia

Urine pHduring acidosis

Determineplasmacortisol

Low Normal

Adrenal Selectiveinsufficiency aldosterone

deficiency

Fig. 5. Classification of RTA in patients with hyperkalemia. Adaptedfrom reference 49.

cation then chloride in their urine when they are acidemic,indicating the presence of defective acidification. Once thediagnosis of an RTA syndrome has been made the urinary aniongap has no further use since it is abnormal in all of them.

As shown in Figures 4 and 5, the next step is to divide thepatients according to the serum potassium, those with a normalor low serum potassium and those in whom it is elevated. Thefirst group (Fig. 4) can be further subdivided into those patientswho can lower the urine pH below 5.5 and those who cannot.When proximal RTA had been excluded, distal RTA can be

Table 3. Patterns of plasma aldosterone and urine pH

Urine pHduring acidemia

Plasmaaldosterone

Aldosterone deficiencyAldosterone resistanceVoltage-dependent distal

RTAAldosterone deficiency

and voltage-dependentdistal RTA combined

<5.5<5.5>5.5

>5.5

LowNormal or highNormal or high

Low

diagnosed with certainty in those whose urine pH is greaterthan 5.5. As mentioned above, the defect in those patientswhose potassium is normal is hypothetically ascribed to adefect in the H-ATPase, while those with hypokalemia arethought to have impairment of the H-K-ATPase. Patients whocan lower urine pH below 5.5 should have proximal RTAexcluded. They then can be given sodium bicarbonate intrave-nously; if the urine pCO2 fails to rise normally, the diagnosis ofrate dependent dRTA can be made.

Patients with hyperkalemia should also be divided on thebasis of urine pH (Fig. 5). In those with a pH below 5.5 a normalplasma cortisol combined with a low aldosterone establishes thediagnosis of selective aldosterone deficiency. Since hyperkale-mia stimulates aldosterone release, plasma aldosterone valuescannot accurately be interpreted without reference to the serumpotassium concentration. When the potassium is normal orelevated, the plasma aldosterone concentration should be atleast three times higher. Thus an aldosterone of 15 ng/ml, whileostensibly normal, is low for a patient with a potassium of 6mEq/liter. If both cortisol and aldosterone and cortisol are low,the patient has Addison's disease. A normal cortisol and a lowaldosterone characterizes selective aldosterone deficiency. Thefinding of a normal or high aldosterone suggests aldosteroneresistance. In patients with hyperkalemia and a urine pHpersistently greater than 5.5, the diagnosis of hyperkalemic orvoltage dependent dRTA is made without further evaluation,but one should look for a specific cause such as obstructiveuropathy or hemoglobin S.

Table 3 shows the pattern of plasma aldosterone and urine pHseen in patients with aldosterone deficiency, aldosterone resis-tance, voltage dependent dRTA, and the combination of aldo-sterone deficiency and RTA. Note that the patient with thecombined disorder is easily distinguished from those with theother disorders. The high urine pH indicates dRTA and the lowaldosterone discloses the steroid deficiency.

Principles of treatmentThe normokalemic and hypokalemic forms of dRTA require

only small amounts of orally administered bicarbonate, or itsequivalent. Because the distal nephron modulates only about100 mEq of acid excretion a day, the maximum amount ofbicarbonate necessary to treat a patient with dRTA is 100 mEqa day. In practice, considerably less is required since most ofthese patients retain the ability to excrete some acid. Childrenare an exception to this rule because of the large alkali require-ment of growing bone. If the defect in hypokalemic patients isa damaged H-K-ATPase, then potassium supplementation mayhave to be continued even after the acidosis has been corrected.

Aldosteroneresistance

726 Kurtzman: Disorders of distal acid{fication

Additional studies of potassium balance in these patients areneeded to carefully examine this issue.

Patients with hyperkalemic (voltage-dependent) dRTAshould first have their underlying disease treated, if one isidentified and is amenable to treatment. Obstructive uropathy isof course the most important example. In general, distal sodiumdelivery should be encouraged by ingestion of generousamounts of salt to stimulate whatever residual capacity for CCTsodium transport remains. Because many patients with thissyndrome have renal and cardiovascular disease, increaseddietary salt can result in edema rather than in increased distaldelivery of sodium, but this problem can be overcome by theaddition of furosemide to the high salt diet. This combinationencourages the distal delivery of salt and stimulates CCT acidand potassium excretion, to the extent that this capacity re-mains intact.

Mineralocorticoid therapy is sometimes useful in patientswith aldosterone deficiency, although increasing distal sodiumdelivery is usually sufficient. Giving steroid without increasingdistal sodium delivery can precipitate heart failure. This com-plication can be avoided with diuretic administration and a diethigh in salt, since salt restriction exacerbates hyperkalemia andmetabolic acidosis. Other forms of treatment, such as potas-sium exchange resin and alkali, may also be necessary.

Acknowledgments

The research by the author described in this paper was supported inpart by a grant from the National Institutes of Health, #RO1-DK-36199.Stephen Norris, M.D., made the illustrations used in Figures 1 to 3.Sandra Sabatini, Ph.D., M.D., and Melvin Laski, M.D., providednumerous and helpful discussions of the theories presented herein.

Reprint requests to Neil A. Kurtzman, M.D., Department of InternalMedicine, Texas Tech University Health Sciences Center, Lubbock,Texas 79430, USA.

References

I. SCHUSTER VL: Bicarbonate reabsorption and secretion in thecortical and outer medullary collecting tubule. Sem Nephrol 10:

139—147, 19902. LASKI ME: Total CO2 flux in isolated tubules during carbonic

anhydrase inhibition: Unique effects of ouabain and amiloride. AmJ Physiol 252:F322—F330, 1987

3. Wioo CS: Active proton secretion and potassium absorption inthe rabbit outer medullary collecting duct. J Gun Invest 84:361—365,

19894. GARG LC, NARANG N: Ouabain-insensitive K-adenosine triphos-

phatase in distal nephron segments of the rabbit. J Clin Invest

81:1204—1208, 19885. WEINER D, HAMM LL: Regulation of intracellular pH in the rabbit

cortical collecting tubule. J Clin Invest 85:274—281, 19906. SELDIN DW, COLEMAN AJ, CARTER N, RECTOR FC JR: The effect

of Na2SO4 on urinary acidification in chronic renal disease. J LabGlitz Med 69:893—903, 1967

7. BATLLE DC: Segmental characterization of defects in collectingtubule acidification. Kidney Int 30:546—554, 1986

8. Purrs RF, LOTSPEICH WD: Bicarbonate and the renal regulation ofacid-base balance. Am J Physiol 147:138—154, 1946

9. HALPERIN ML, GOLDSTEIN MB, HAlO AJ, STINEBAUGH BJ: Stud-ies on the pathogenesis of type 1 (distal) renal tubular acidosis asrevealed by the urinary PCO2 tensions. J Clin Invest 53:669—677,1974

10. ARRUDA JAL, NASCIMENTO L, KUMAR S, KURTZMAN NA: Fac-tors influencing the formation of urinary carbon dioxide tension.Kidney Int 11:307—317, 1977

11. DUBOSE TD JR: Hydrogen ion secretion by the collecting duct as adeterminant of the urine to blood PCO2 gradient in alkaline urine. JClin Invest 69:145—156, 1982

12. KURTZMAN NA: Acquired distal renal tubular acidosis. Kidney Int

24:807—819, 198313. BATLLE DC, HIzON M, COHEN E, GUTTERMAN C, GUPTA R: The

use of the urine anion gap in the diagnosis of hyperchloremicmetabolic acidosis. N EngI J Med 3 18:594-599, 1988

14. LIGHTw00D R: Communication No. 1. Arch Dis Child 10:205—206,1935

15. BUTLER AM, WILSON JL, FARBER 5: Dehydration and acidosiswith calcification at renal tubules. J Pediatr 8:489—499, 1936

16. ALBRIGHT F, BURNETT CH, PARSON W: Osteomalacia and laterickets. The various etiologies met in the United States withemphasis on that resulting from a specific form of renal acidosis, thetherapeutic indications for each etiological subgroup, and therelationship between osteomalacia and Milkman's syndrome. Med-icine 25:399—479, 1946

17. SELDIN DW, WILSON JD: Renal tubular acidosis, in MetabolicBasis of Inherited Diseases (3rd ed), edited by STANBURY JB,WYNGARDEN JB, FREDRICKSON D, New York, McGraw Hill BookCo, 1972, pp 1548—1566

18. DUBOSE TD JR, CAFLISCH CR: Validation of the differences inurine and blood CO2 tension during bicarbonate loading as an indexof distal nephron acidification in experimental models of distal renaltubular acidosis. J Clin Invest 75:1116—1123, 1985

19. STEINMETZ PR, LAWSON LR: Defect in acidification induced invitro by amphotericin-B. J Clin Invest 49:596—601, 1970

20. BATLLE DC, SEHY iT, RO5EMAN MK, ET AL: Clinical and patho-physiologic spectrum of acquired distal renal tubular acidosis.Kidney mt 20:389—396, 1981

21. THIRAKOMEN K, KosLov N, ARzwDAJAL, KURTZMAN NA: Renalhydrogen ion secretion following the release of unilateral ureteralobstruction. Am J Physiol 23 1:1233—1239, 1976

22. MEHTA PK, SODHI B, ARRUDA JAL, KURTZMAN NA: Interactionof amiloride and lithium chloride on distal acidification. J Lab ClinMed 93:983—994, 1979

23. ARRUDA JAL, SUBBARAYUDU K, DYTKO G, MOLA R, KURTZMANNA: Voltage dependent distal acidification defect induced byamilonde. J Lab Gun Med 95:407—416, 1980

24. LASKI ME, KURTZMAN NA: Characterization of acidification in thecortical and medullary collecting tubule of the rabbit. J Gun Invest72:2050—2059, 1983

25. KURTZMAN NA: "Short circuit" renal tubular acidosis. J Lab ClinMed 95:633—636, 1980

26. BATLLE DC, ARRUDA JAL, KURTZMAN NA: Hyperkalemic distalrenal tubular acidosis associated with obstructive uropathy. NEnglJMed3O4:373, 1981

27. BATLLE DC, ITSARAYOUNGYUEN K, ARRUDA JAL, KURTZMANNA: Hyperkalemic hyperchloremic metabolic acidosis in sickle cellhemoglobinopathies. Am J Med 72:188—192, 1982

28. LA5KI ME, KURTZMAN NA: Site of the acidification defect in theperfused post-obstructed collecting tubule. Miner Electrol Metab15:195—200, 1989

29. SABATINI S. KURTZMAN NA: Enzyme activity in obstructiveuropathy: The biochemical basis for salt wastage and the acidifica-tion defect. Kidney mt 37:79—84, 1990

30. DAPHNIS E, KURTZMAN NA, SABATINI S: On the mechanism oflithium-induced renal tubular acidosis. (abstract) Kidney Int 37:534A, 1990

31. NASCIMENTO L, RADEMACHER DR, HAMBURGER R, ARRUDA JAL,KURTZMAN NA: On the mechanism of lithium induced renaltubular acidosis. J Lab Clin Med 89:455—462, 1977

32. ARRUDA JAL, DYTKO G, MOLA R, KURTZMAN NA: On themechanism of lithium-induced distal renal tubular acidosis: Studiesin the turtle bladder. Kidney mt 17:196—204, 1980

33. BATLLE DC, GAvIRIA M, GRUP M, ARRUDA JAL, KURTZMAN NA:Distal nephron fuction in patients receiving chronic lithium ther-apy. Kidney mt 21:477—485, 1982

34. BATLLE DC, MOZE5 MF, MANALIGOD J, ARRUDA JAL, KURTZ-MAN NA: The pathogenesis of hyperchloremic metabolic acidosisassociated with kidney transplantation. Am J Med 70:786—796, 1981

35. BATLLE DC, GRUPP M, GAVIRIA M, KURTZMAN NA: Distal renal

Kurtzman: Disorders of distal acidification 727

tubular acidosis with intact capacity to lower urinary pH. Am JMed 72:751, 1982

36. HUDSON JB, CHOBANIAN AV, RELMAN AS: Hypoaldosteronism, aclinical study of a patient with an isolated adrenal mineralocorticoiddeficiency resulting in hyperkalemia and Stokes-Adams attacks. NEngl J Med 257:530—536, 1957

37. BATLLE DC: Hyperkalemic hyperchloremic metabolic acidosisassociated with selective aldosterone deficiency and distal renaltubular acidosis. Sem Nephrol 1:260—273, 1981

38. SHAMBELAN M, SEBASTIAN A, BIGLIERI EG: Prevalence, patho-genesis, and functional significance of aldosterone deficiency inhyperkalemic patients with chronic renal insufficiency. Kidney mt17:89—101, 1980

39. SZYLMAN P, BETTER OS, CHAIMOWITZ C: Role of hyperkalemia inthe metabolic acidosis of isolated hypoaldosteronism. N Engi JMed 294:361—365, 1976

40. ARRUDA JAL, BATLLE DC, SEHY JT, R0SEMAN MK, BARA-NOWSKI RL, KURTZMAN NA: Hyperkalemia in renal insufficiency:Role of aldosterone deficiency and renal tubular unresponsiveness.AmJNephrol 1:160—167,1981

41. DITELLA PJ, SODHI B, MCCREARY J, ARRUDA JAL, KURTZMANNA: Mechanism of the metabolic acidosis of selective mineralocor-ticoid deficiency. Kidney mt 14:466—477, 1978

42. HULTER HN, ILNICKI LP, HARBOTTLE JA, SEBASTAIN A: Impairedrenal H secretion and NH3 production in mineralcorticoid-defi-

cient glucocorticoid-replete dogs. Am J Physiol 232:F136—F146,1977

43. Muiis SK, NASCIMENTO L, RADEMACHER DR, WILSON A,KURTZMAN NA: Intact ability to lower urine pH in non-acidoticadrenalectomized rats. Miner Electrol Metab 12:107—112, 1986

44. DONNEL GN, LITMAN N, ROLDAN M: Pseudohypoadrenalocorti-cism. Renal sodium loss, hyponatremia, and hyperkalemia due to arenal tubular insensitivity to mineralocorticoids. Am J Dis Child97:813—817, 1959

45. STANBURY SW, MAHLER RF: Salt-wasting renal disease. Metabolicobservations on a patient with "salt-losing nephntis." QJ Med28:425—447, 1959

46. WALKER WG, JosT U, JOHNSON JR, KOWARSKI A: Metabolicobservations on salt wasting in a patient with renal disease. Am JMed 39:505—5 19, 1965

47. ARNOLD JE, HEALY JK: Hyperkalemia, hypertension and systemicacidosis without renal failure associated with a tubular defect inpotassium excretion. Am J Med 47:461—472, 1969

48. GORDON RD, GEDDES RA, OAWSEY CGK: Hypertension andsevere hyperkalemia associated with suppression of renin andaldosterone and completely reversed by dietary sodium restriction.Aust Ann Med 4:287—294, 1970

49. KURTZMAN NA: Renal tubular acidosis: A constellation of syn-dromes. Hosp Prac 22:173—188, 1987