63I

DIURETICSA Review

By H. G. LLOYD-THOMAS, M.A., M.B., M.R.C.P.Senior Registrar, The London Hospital

All substances that increase urine flow may beclassed as diuretics, and in this sense water itselfis the foremost diuretic. As the main purpose ofusing diuretics is to promote the excretion ofwater, it is customary to restrict the term' diuretic'to substances which induce a net loss of waterfrom the body. It is also usual to exclude sub-stances such as digitalis in which the main actionof the drug is on the heart.The chemist has provided the clinician with a

large number of substances which have diureticactivity. The pharmacological actions of severalof these will be considered.

Until recent years the mercurials have enjoyedan undisputed pre-eminence among the diuretics.For a long time there has been, however, a realneed for an effective and non-toxic oral diureticthat would free the patient from the pain ofinjected mersalyl and the inconvenience of injec-tions, and which would provide a more uniformmaintenance of an oedema-free state. Manypatients feel weak and unwell for 24 to 48 hoursfollowing a vigorous diuretic response to mersalyl.The search for such a drug continues. Amongstseveral diuretics suitable for oral administrationdeveloped since 1950, chlorothiazide (' Saluric,'' Chlotride,' ' Diuril') has proved to be the mostserious rival to the organic mercurials.

In the past, the use of diuretics was mainlyrestricted to the patient with congestive heartfailure or oedema due to hepatic cirrhosis. Sinceorally-effective agents have been made available,the indications for the use of diuretics have becomebroader and include conditions such as pre-menstrual fluid retention, and pregnancy oedema.

The Organic Mercurial DiureticsMersalyl ('Salyrgan'), introduced in I924, is

one of the most commonly used preparations. Itshould be given by intramuscular injection, theusual dose being 2 ml. Intravenous injection maylead to immediate toxic cardiac reaction withventricular fibrillation and sudden death (DeGraff et al., 1942).

Oral preparations such as chlormeredrin tendto cause alimentary side-effects and they producea smaller diuretic response for the same totaldosage of mercury ion.Mercaptomerin (' Thiomerin ') has the advantage

that it can be given subcutaneously and can,therefore, be self-administered, but pain, eccy-mosis and fibrous nodules may occur.

The primary action of the organic mercurials isto depress tubular reabsorption of certain ions.This effect can be blocked by some mercaptans.As all the transport systems in the renal tubulesare not inhibited by mercury, the urine secreted inresponse to a mercurial diuretic has a certainpattern which will be discussed below. As aresult of a proportionately greater loss of chloridethan bicarbonate from the extracellular fluid thereis a tendency for hypochloraemic alkalosis todevelop.The diuretic response may be potentiated by the

preceding administration of ammonium chloride,which also tends to reduce excessive chloridedepletion. Factors which may be responsible forunresponsiveness to mercurial diuretics, eventhough oedema remains, include hypochloraemia,a low glomerular-filtration rate, and the severityof the patient's condition (Stock et al., 1951).The excretion of the injected mercury is fairly

rapid: 50 per cent. can be recovered in threehours, 95 per cent. within 24 hours. Measurableamounts are excreted on the second and thirddays and, therefore, cumulative effects may occurfrom injections given daily (Burch et al., 1950).If there is impaired renal function the excretionmay be delayed.Contraindications

Mercurial diuretics are absolutely contra-indicated when there is an active nephritis, andwhen there is renal insufficiency with elevation ofthe blood urea above 60 mg. per Ioo ml. becausethe drug may aggravate the existing renal lesion(Preedy and Russell, 1953), and there may bedelayed excretion of the mercury. The presence

I

Protected by copyright.

on March 5, 2020 by guest.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.35.409.631 on 1 N

ovember 1959. D

ownloaded from

POSTGRADUATE MEDICAL JOURNAL

of albuminuria with casts and red cells demandscaution, - although these may, of course,- resultfrom congestive changes in the kidney. Whenthere is enlargement of the prostate,; dfrinary'retention may follow a profuse diuretic response.

Untoward reactions occur infrequently. Sudden ,death may follow intravenous injection. Im-mediate non-fatal reactions may consist of flushing,chills, rigors, fever, urticaria, pruritis, skin erup-tions. Dyspnoea; anxiety, palpitation, pallor andfaintness may occur and should not be ignored. Adifferent preparation of mercurial diuretic mayavoid such reactions. Systemic mercury poisoningwith dermatitis, stomatitis and colitis usuallyoccurs when the diuretic response has been poor,particularly if the frequency of the injections hasbeen injudiciously increased on this account.Fatal anuria from tubular necrosis may occur insuch circumstances. Neutropenia and agranulo-cytosis has been reported (Silverman and Worthen,1952). Exfoliative dermatitis has been reportedafter mercaptomerin ('Thiomerin') (MacHaffieet al., 1958).

Secondary toxic effects are due to depletion ofextracellular electrolyte. The low-salt syndromemay result and, unless corrected, may progress tooliguria and nitrogen retention. Digitalis in-toxication may rarely occur after massive diuresis.

ChlorothiazideChlorothiazide, synthesized by Novello and

Sprague in I957, is designated chemically as6-chloro-7-sulfamyl-I, 2, 4-benzo-thiadiazine-I, i-dioxide. It contains an unusual fused-ringsystem wherein sulphur is in its highest oxidationstate, as well as a sulphamyl group attached to abenzenoid ring. The compound possesses a highorder of activity as an inhibitor of carbonicanhydrase in vitro (Beyer et al., 1957).

Oral absorption is excellent as measured by theprompt electrolyte and water excretion responses,and excretion occurs rapidly via the kidney.Although it has carbonic.anhydrase inhibitor

activity in vitro, its diuretic action in man islargely due to increased excretion of chloride ionswith an almost equal loss of sodium. In itschloruretic action it resembles closely the organicmercurial compounds (Pitts and Sartorius, I950).Increase of bicarbonate excretion may occur in theearlier hours of the action of the drug, but com-pared with acetazolamide. the effect is neithermarked nor constant (Ford and Spurr, 1957;Bayliss et al., i958a). The excretion of potassiumin the urine, which may be sufficiently great toproduce a state of potassium depletion, mayibe dueto carbonic anhydrase inhibition..

It is most unusual for mercurial diuretics to

induce hypokalaemia. As mercurials may inducea greater loss of chloride' than of sodium, hypo-chloraemia may occur and make the subject un-

,rsponsive to further mercurials. Such refrac-toriness is unlikely to be induced by chloro-

?thiazide, as this drug produces almost equimolarloss of sodium and chloride. The action of chloro-thiazide is likely to be maintained and because ofthis it is more likely to produce potassium de-pletion.

Chlorothiazide will frequently effectively relieveoedema'and the pulmonary congestion due toheart failure (Moyer et al., I957; Bayliss et al.,1958a; Bayliss et al., I958b; Slater and Nabarro,1958; Davies and Evans, 1958; Dinon etal., I958).It is of value in causing the reduction or disap-pearance of oedema in nephrosis and chronicglomerulo-nephritis (Slater and Nabarro, 1958;Schreiner and. Bloomer, I957). In patients withcirrhosis of the liver with ascites, a good diureticresponse was obtained in nine out of 13 cases(Read et al., 1958), poor responses being obtainedin those with an initial urinary output of sodiumof less than one mEq. daily. Finnerty et al. (g958)found the drug to be useful in the treatment of theoedema and of the hypertension present in pre-eclamptic toxaemia. Premenstrual oedema maybe completely relieved (Bayliss et al., I958b;Dinon, 1958).Dosage of chlorothiazide must be determined

for each patient individually: 2 g. produces amaximum diuresis and this is approximately equalto the effect of 2 ml. of mercurial (Davies and Evans,1958). It is rarely necessary to give 2 g. daily, andsuch a dose is probably better not continued forlonger than five days without careful electrolytecontrol. Many patients do well on'2 g. per daygiven on alternate days or on two days a week.Even smaller doses may be sufficient. Disturbanceof the patient's sleep may be avoided if the drug isgiven in the early part of the day.

Patients who are refractory to mercurial injec-tions respond well to chlorothiazide and thereverse effect may also occur (Bayliss et al.,I958b; Davies and Evans, I958). The use ofammonium chloride with chlorothiazide may leadto a better diuresis (Watson et al., 1958).Untoward effects are uncommon. Direct toxic

effects that may be attributed to the drug itself areextremely rare. Drerup (I958) reported a patientin whom jaundice occurred which he thoughtmight have been related to chlorothiazide. Dinonet al. (1958) reported neutropenia and thrombo-cytopenia in one patient receiving chlorothiazideand a nearly fatal allergic reaction in a secondpatient.The most important side-effect of chlorothiazide

is due to its effect in causing potassiumlss leading

632 November 1959

Protected by copyright.

on March 5, 2020 by guest.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.35.409.631 on 1 N

ovember 1959. D

ownloaded from

LLOYD-THOMAS: Diuretics

to potassinmidepletion which need not be accom-panied.!by symptoms (Goodkind et al., 1958;Laragh.^t:. al., 1958). Hypokalaemia is of greatimportance in patients with cirrhosis of the liverand. scites' as it may precipitate hepatic coma(Read et,, , 1958; Mackie et al., 1958). In-creased' sensitivity .to the toxic effects of digitalisoccurs when there is potassium depletion. Hypo-kalaemia usually occurs in the seriously ill patients,especially those with a low dietetic intake ofpotassium., It is best avoided by using inter-mittent dosage of chlorothiazide and by sup-plementing the, intake of potassium, using a mix-ture rather than tablets, which may be passed un-changed; in the stool. A suitable mixture isas follows:-

Potassium chloride I g.Syrup simplex 2 ml.Alcohol, 95 per cent. 2 ml.Raspberry flavouring qs.Chloroform water to 15 ml.

This may be given two to six times per dayas needed.

It may, however, be difficult to control hypo-kalaemia while continuing therapy with the drug(Slater and Nabarro, I958). Prolonged hypo-kalaemia,should be avoided because of the risk ofrenal (Follis, R. H., 1948; Perkins et al., 1950;Conn et al.. 1956) and cardiac injury (McAllen,I955).Care should be exercised when this.preparation

is to. be given to patients already on hypotensivetherapy because of the marked potentiating effectchlorothiazide may have on many of the hypoten-sive drugs now in use.

Hydroch1grothiazideHydrochlorothiazide, (' Hydrosaluric, Esidrex')

is a new preparation which is undergoing trial atpresent. The electrolyte excretionpattern followingits use is similar to that produced by chlorothiazidebut the.. ffect,on potassium and bicarbonate ex-cretion is less.. The effective dose is smaller, ashydrochlorothiazide appears to be five to ten timesmore potent on a milligram for milligram basis.

Drugs of the Amino-Uracil Group.f the active diuretics in the amino-uracil

series of, drugs, aminometradine (' Mictine')(Kattus et al., 19.52; Platts and Hanley, I956;Hanley ,nd Platts; I956) and amisometradine(' Rolicton') have been the most widely investi-gated. They are believed to act by direct inhibi-tion of tubular reabsorption of sodium. Amino-metradine given orally is generally found to be atleast 50 per pent. as effective as.mersalyl by injec-tion, but effective doses cause epigastric discom-fort, nausea, vomiting or diarrhoea in a consider-

able number of subjects. Headache, tinnitus andalbuminuria are less common side-effects.

Jose and Wood (I958) tested. the effect ofamisometradine in 2o patients with heart failureusing a dose of 400 mg. three times, daily for twoconsecutive days. They found that the diureticpotency of the drug was approximately 40 percent. that of 2 ml. of mersalyl given intramus-cularly. Apart from nausea in one patient andtinnitus in another, no toxic effects were observedin this short-term study, Diarrhoea was reportedin two patients of a series of 47 cases reported bySettle (I957). He found that there was a decreaseof 88 per.cent. in the number of mercurial injecetions required in patients who were maintained onregular therapy with amisometradine.The author has; observed several cases with

chronic cardiac oedema in whom neither mer-curial diuretics (2 ml.) nor aminometradine(200 mg. q.d.s.) alone produced a diuretic response,but when these drugs were administered on thesame day, a massive diuresis occurred with loss ofoedema. The loss of potassium in the urine isusually little increased by this group of drugs, andthey may, therefore, be of use particularly in thosepatients in whom.the use of chlorothiazide causesexcessive potassium wastage. The slight tendencytoward depression of the serum sodium, potassiumand chloride, which occurs with the daily ad-ministration of. aminometradine, never becomessignificant, since the drug ceases to be effective oncontinued daily use (Ford et al., 1958). Thisproblem of absence of effectiveness (in spite of thepresence of oedema) on repeated daily use,becauseof drug,tolerance is a problem common to thisgroup of drugs and to the carbonic anhydraseinhibitors..,

Drugs which inhibit Carbonic AnhydraseAcetazolamide (' Diamox '), designated chemi-

cally as 2- acetylamino-i, 3, 4- thiadiazole-5-sul-phonamide,. a carbonic anhydrase inhibitor, pro-duces a diuresis by increasing the urinary loss ofbicarbonate, which, is.accompanied by increasedexcretion of sodium, potassium and water. Itsaction is not as rapid, nor as dramatic, as that ofanorganic mercurial. After a few days' continuedadministration the drug loses its,natriuretic effectand its action is self-limiting because a metabolicacidosis develops and the amount of bicarbonatefiltered by the glomeruli is reduced (Counihanet al., I955). Hanley.and Platts (I956) found itless effective than Mersalyl. Relman et al. (I954)concluded that in patients with severe congestiveheart failure, it was of little or no value. Goldet al. (1958) found that acetazolamide in a dose ofI25 to i,ooo mg. had a diuretic effect approximatelyequal to 0.5 mi. of injected meralluride, while

XavemntbT3t,j:9.59 633P

rotected by copyright. on M

arch 5, 2020 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.35.409.631 on 1 Novem

ber 1959. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

ethoxzolamide (6-ethoxy-2-benzothiazole-sulphonamide) had a diuretic effect equal to0.7 ml. of meralluride. In a dose of 62.5 to 250 mg.it gave rise to diarrhoea, colic and vomiting.

If acetazolamide is to be used as a diuretic it isbest to use it as a single dose of 250 to 500 mg.every other day, perhaps combining it with anintramuscular mercurial diuretic given once ortwice a week. Oedema associated with renaldamage is a contraindication, since the kidney maybe unable to compensate sufficiently for themetabolic acidosis. The drug should not be usedwhen the oedematous state is associated withpotassium-losing conditions such as steroidoedema, cirrhosis or nutritional oedema. Websterand Davidson (I956) report on the precipitation ofhepatic coma by acetazolamide and this may bepartly related to potassium deficiency.Untoward EffectsNumerous side-effects from the use of acetazola-

mide have been reported. The most common side-effects are paraesthesiae and drowsiness (Friedbergand Halpern, I952). Fever, leucopenia, skinrashes, may occur. Other manifestations of sul-phonamide reaction, such as haemolytic anaemia,bone-marrow depression with agranulocytosis(Pearson et al., 1955) and thrombocytopenia(Reisner and Morgan, 1956), as well as crystalluriawith renal colic and anuria (Yates-Bell, 1958;Glushien and Fisher, 1956; Persky et al., 1956),may arise.

XanthinesWhen given in large doses these substances

result in increased cardiac output, renal bloodflow and glomerular filtration rate. They are alsocapable of depressing renal tubular reabsorptionof electrolyte (Walker et al., 1937).The xanthines are little used as diuretics at

present as there is a wide choice of more effectivepreparations. Theophylline is included in mostpreparations of organic mercurial injections withthe object of increasing the rate of absorption ofthe mercurial from the site of injection.Acid-forming SaltsThe most widely used of these substances is

ammonium chloride which is a combination of alabile cation and a fixed anion. The ammoniumion is converted to urea which leaves excesschloride ion. This displaces bicarbonate ionwhich is converted to CO2 and excreted. Thetotal electrolyte concentration of the extracellularfluid is unchanged as a labile anion and a labilecation disappear, but the chloride concentrationin the extracellular fluid tends to rise and, as aresult, the chloride load delivered to the renal

tubule is increased. That amount of chloridewhich escapes reabsorption in the tubules isexcreted together with an equivalent amount ofcation (mostly sodium) and water.The main use of ammonium chloride is in com-

bination with mercurial diuretics (and possiblychlorothiazide). By increasing the concentrationof extracellular chloride the action of the mercurialis potentiated and the tendency to the productionof hypochloraemia counteracted.Evans and Paxon (I94I) found that the best way

to use ammonium chloride was to administer30 gr. in enteric-coated tablets two hours beforethe injection of mercurial.

Acid-forming salts are contraindicated if thereis impairment of renal function because of thedanger of causing uncompensated acidosis.

Osmotic DiureticsTwo only are discussed.

Potassium SaltsThe renal tubular secretion of potassium

markedly depresses H+-Na+ exchange on therenal tubule. This results in the excretion of analkaline urine containing sodium salts, and a netloss of extracellular electrolyte and water canthereby result (Berliner et al., 1951).They are generally not very effective as diuretics,

and the maximum benefit is achieved when a dietlow in sodium chloride is used. Their use iscontraindicated when renal failure is present.UreaThis substance is now little used as a diuretic.

It has to be given in large doses of 20 g. two to fivetimes per day. Its use is contraindicated whenthere is impairment of renal function.

Steroidal Antagonists of Aldosterone(Spirolactone)

Aldosterone has been implicated in the patho-genesis of abnormal sodium and water accumula-tion which occurs in oedematous states such ascongestive heart failure, cirrhosis of the liver withascites and nephrosis. Many patients with theseconditions excrete abnormally large amounts ofaldosterone in the urine (Axelrad et al., 1955;Chart and Shipley, 1953; Duncan et al., I956;Leutscher and Johnson, 1954; Wolff andKoczorek, 1955; Wolff et al., I958; Leutscheret al., I954). The excessive secretion of aldos-terone may indicate over-production of thissteroid which may be at least in part responsiblefor sodium retention. Measures that combat theeffects of excess aldosterone may, in those caseswhere this mechanism is operating, be of use inrelieving the oedematous state.

634 November I959P

rotected by copyright. on M

arch 5, 2020 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.35.409.631 on 1 Novem

ber 1959. Dow

nloaded from

POSTGRADUATE MEDICAL JOURNAL

Recently, two synthetic steroids, 3-(3-oxo-I7P-hydroxy-4-androsten-I7a-yl) propionic acidy-lactone (SC 5233) and its i9-nor analogue (SC8o09), have been shown to possess natriureticactivity consequent upon their anti-aldosteroneactivity (Cella and Kawaga, 1957; Kawaga et al.,1957) and therefore manifested only in the presenceof sodium-retaining steroids, endogenous or exo-genous (Liddle, G. W., 1957). In patients withAddison's disease, maintained only on a high saltintake alone, SC 8I09 had no effect on the excretionof sodium, potassium, chloride, phosphate, am-monia or titratable acid (Liddle, G. W., I957, perKerr et al., 1958), but, when given'during treat-ment with DOCA increased excretion of water,sodium and chloride and decreased; excretion ofpotassium, phosphate, ammonia and titratable acidoccurred. Normal subjects on a high sodiumintake do not show significant responses tospirolactones, but when a state of high aldosteroneoutput is induced in the same subjects by restric-tion of sodium intake, significant responses followthe administration of the spirolactones. Nochanges in the output of aldosterone occur inshort-term studies, but continued administrationof the spirolactones may be associated with a risein aldosterone excretion. Salassa et al. (1958), ina study on a patient with primary aldosteronism,found that SC 8o09 caused a marked effect on theexcretion of sodium when a large amount ofaldosterone was being excreted; it caused anegligible effect--when, following removal of anadrenal cortical adenoma, only a trace of aldos-terone was being excreted. These findings suggestthat the spirolactopes act as antagonists to aldos-teroiinE at 'en'd-orgdin' evel (the renal tubule).Liddle (I957) found a consistent natriuresis inseven patients with congestive.heart failure treatedwith spirolactone. Kerr et al. (1958) obtained asodiuamand water diuresis in two cases of cirrhosisof the liver with ascites and failed to do so in athird patient. treated with SC 8109.- The negligiblepotassium loss is an important feature of the diureticresponse during treatment wnitlSC 8o09. Diureticswhich cause a heavy loss of.potassium should beavoided in patients with cirrhosis of the liver whohave a tendency towards hepatic coma. The bestmethod for the administration of these new drugshas yet to be defined. Oral-and-intramuscularpreparations are obtainable. They are not com-mercially available because of prohibitive cost.

Urinary Electrolyte Excretion PatternsPrecse knowledge concerning the mechanism of

action of the main diuretics used today is not com-plete. The principal action is the inhibition ofrenal tubular reabsorption of sodium. Evidencethat the fundamental method by which this is

accomplished differs among various diuretic drugsis afforded by study of the urinary electrolyteexcretion that follows their use. Ford (1958)studied patients with incipient heart failure andobtained the following results.The urinary electrolyte patterns which follow

the oral or intramuscular administration of theorganic mercurials are similar. Sodium andchloride excretion are increased for 12 to I8 hourswith the maximum effect occurring during thefirst two to six hours. Potassium excretion showsno significant alteration. Ammonia excretion isslightly depressed during the period of maximumsodium and chloride loss, but is elevated after theperiod of greatest sodium loss. During the periodof maximum chloride loss, there is suppression ofbicarbonate loss which returns to normal or greaterthan normal when the chloride loss begins to falloff, this action lasting for 12 to I8 hours. Theresponse of phosphate excretion is similar to thatof bicarbonate.The urinary electrolyte pattern which follows

the oral administration of drugs of the amino-uracilgroup (amisometridine and aminometridine) showsan increase in sodium and chlori4e excretion withlittle o:f: iffit in it ; i9 bi'rbonateexcrtt.i :;hset of aci^'of,,these drugsoccu'rsiWin- tw ot fur ours ai3c3ahtinues forappoxii a hburs:' Thfere ~ri significantchgs imnhosphae or,rimniiaexcretion.'i

i^Thn3g n theaditmtr fwtiSoi t̂A #'te ^ thin twohours and continues for approximately I2 hours.There is an increase in the excretion of sodiumand chloride, with,an increase in the excretion ofpotassium amounting' to about'one half that ofsodium. Ammonia excretion is not changedduring the period of sodium loss but shows someincrease after I2 hours. The increase in excretionof bicarbonate is about a quarter that of chloride.Phosphate excretion shows no significant change.The urinary electrolyte pattern following the

administration of acetazolamide shows an increasein the excretion of sodium with an almost equaleffect on potassium and bicarbonate. Chlorideexcretion is depressed during the period ofmaximum loss of bicarbonate but increases as thebicarbonate: excretion diminishes.' There is nosignificant change in ammonia excretion nor inphosphate excretion, The onset of action of thedrug occurs within six hours and lasts for six totwelve hours.The urinary electrolyte pattern following the

administration of spirolactones has already beenindicated.When assessment of a diuretic response is nade,

it is essential to consider the formidable difficultiesinvolved (Spencer and Lloyd-Thomas, I953). The

6,36 November 1959P

rotected by copyright. on M

arch 5, 2020 by guest.http://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.35.409.631 on 1 Novem

ber 1959. Dow

nloaded from

November 1959 . LOYD-THOMAS: Diuretics 637

test subjects differ in the nature of their disease, in'the causation of their oedema, in renal functionand in many other variable factors important in thedifferential distribution of their body fluids, suchas the-concentration of serum albumin and plasmasodium. Not only do patients differ among them-selves, but there. is a, considerable daily variationof the renal excretion of water and electrolytes inthe same patient which is still the case when theintake is constant, within the limits of a standardlow-sodium diet.

The Comparative Potency of Diuretic DrugsAscribing a ':potency' of one to an organic

mercurial given by intramuscular injection, thefollowing degrees of potency may be ascribed tochlorothiazide, aminometridine, and acetazolamiderespectively, 0.8, 0.7 and 0.25 (Ford, I958). Thesevalues are based on the results of single doses anddo not, therefore, accurately reflect the usefulnessof the drugs. Both acetazolamide and amino-metridine lose their efficiency with continuedadministration and this is a very important factorwhich seriously limits the value of these drugs astherapeutic agents.Additional Measures. for the Mobilization ofOedema

Restriction of sodium intake is still essential ina large number of oedematous patients.A fundamental problem is that concerned with

water intake in patients with oedema'. It is nowgenerally agreed that the intake of' extracellularelectrolytes rather than water is important in thisregard, and that if sodium salts are 'sufficientlyrestricted nothing is gained by the simultaneousrestriction' of water. If, however, the patient willnot accept adequate restriction of sodium intake,it is probably beneficial to restrict the intake ofwater.

In certain patients, who show no response tothe diuretic agents described above, it may benecessary to resort to drainage of the legs bymeans of multiple punctures or by the use ofSouthey's tubes under systemic antibiotic cover.After a satisfactory 'drainage, it may often be foundthat the patient will again respond to diuretics.Occasionally a diuresis will follow venesection anda similar event may occur after a period of fever.The use of cation-exchange resins is now rarely

required in the treatment of cardiac oedema. Theiruse in treatment had been 'described by Spencerand Lloyd-Thomas (i954) and their place in thetreatment of the nephrotic syndrome evaluated byRosenheim and Spencer (1956). The place ofcortisone and its analogues in the treatment of thenephrotic syndrome has been reviewed by Smart(I958). Ross and Ross (i957) recorded deaths

from cellulitis in two nephrotic patients treatedwith cortisone.Rad6 and Blumenfeld (1958) record the use of

ACTH to restoYe responsiveness to mercurialdiuretics,

It may in certain patients be necessary to use acombination of diuretic agents in order to obtaina satisfactory diuretic response. The use ofammonium chloride to increase the effect ofmercurials and to reduce the incidence of hypo-chloraemia in long-term therapy is well known.The use of potassium salts instead of ammoniumchloride may be more effective occasionally.Mercurials, when given together with chloro-thiazide, may succeed in producing diuresis whenneither drug alone is effective. The successful useof mercurials together with aminometridine hasbeen referred to.

Failure to obtain satisfactory mobilization ofoedema fluid may be due to electrolyte derange-ments which should be corrected as far as possible.The use of salt-poor albumin (Thorn et al.,1945; Patek et al., 1948) and of the plasma-expanders such as dextran in hyperoncotic solutionto increase the colloid osmotic pressure of theplasma is of value on occasions. Space does notallow of further consideration of these substances.Choice of Diuretic AgentThe majority of patients who require diuretic

,therapy will obtain satisfactory results from the useof intramuscular mercurials which are well triedand relatively free from toxic effects. They arestill the standard to which other diuretics are com-pared. Where ease of administration is required,the most potent available oral diuretic should beused. At present this drug is chlorothiazide.The regular use of a small intermittent dose of

schlorothiazide or other oral diuretic may oftenallow the patient to ingest more salt and morehighly salted foods than he would otherwise beable to do. The psychological,advantages of beingallowed a more normal diet are obvious. Manypatients, however, will have to continue withmoderate or even severe restriction of sodiumintake.The clinician should bear in mind the pharma-cological properties of the various diuretics avail-

able, so as to make an intelligent choice of theproper drug or combination of'drugs to fulfil thetherapeutic need of the particular patient, andavoid untoward effects.

BIBLIOGRAPHYAXELRAD, B. J. CATES, J. E, JOHNSON, B. B., andLEUTSCHER, J. A. (x9ss), Brit. med. J., i, 196.BAYLISS, R. I. S MARRACK, D., REES, J. R., and ZILVA,J. F. (i958a), Aim. N. Y. Acad. Sci.; 7I, 442.BAYLISS, R. I. S., MARRACK, D. PIRIIS, J., REES, J. R., andZILVA, J. F. (1958b), Lancet, i, I20.

Protected by copyright.

on March 5, 2020 by guest.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.35.409.631 on 1 N

ovember 1959. D

ownloaded from

638 POSTGRADUATE MEDICAL JOURNAL November x959

BERLINER, R. W., KENNEDY, T. J., and ORLOFF, J. (x95z),Amer. J. Med., x , 274.

BEYER, K. H., BAER, J. E., RUSSO, H. F., and HAIMBACH,A. S. (1957), Fed. Proc., 16, 282.

BURCH, G., THORPE, R., THREEFOOT, S., KELLY, F. J.,and SVEDBORG, A. (1950),J. cln. Invest., 29, 1131.

CELLA, J. A., and KAWAGA, C. M. (i957), y. Amer. chem. Soc.79, 4808.

CHART, J. J., and SHIPLEY, E. S. (1953), . din. Invest., 32, S60.CONN, J. W., and JOHNSTONE, R. D. (x956), Amer. .. cin

Nutrit., 4, 523.COUNIHAN, T. B., EVANS, B. M., and MILNE, M. D. (z955),

Clin. Sd., 13, 585.DAVIES, D. W., and EVANS, B. (I958), Brit. med. ., i, 967.DE GRAFF, A. C., and LEHMAN, R. A. (x942), J. Amer. med.

Ass., 119, 998.DINON, L. R., KIM, Y. S., and VANDER VEER, J. B. (x958),

Amer . ed. Scd., 236, 533.DRERUP, A. L., ALEXANDER, W. A. LUMB, G. D.,

CUMMINS, A. J., and CLARK, G. M. (x958), New Engl. J.Med., 259, 534.

DUNCAN, L. E., LIDDLE, G. W., and BARTTER, F. C. (1956),7. cin. Invest., 35, 1299.

EVANS, W., and PAXON, T. (i94i), Brit. Heart y., 3, 37.FINNERTY, F. A., Jr., BUCHOLTZ, J. H., and TUCKMAN, J.(I958), y. Amer. med. Ass., x66, 141.FORD, R. V., and SPURR, C. L. (i957), Amer. Y. Med., 22, 965.FORD, R. V., ROCHELLE, J. B. III, HANDLEY, C. A.,

MOYER, J. H., and SPURR, C. L. (x958), J. Amer. med. Ass.,x66, 129.

FOLLIS, R. H. Jr. (I948), 'The Pathology of Nutritional Disease.'Charles C. Thomas, Springfield.

FRIEDBERG, C. K., andHALPERN, M. (1952), Fed. Proc., Ix, 49.GLUSHIEN, A. S., and FISHER, E. R. (956), Ibid., I60, 204.GOLD, H., GREINER, T. H., WARSHAW, L., KWIT, N. T.,

and GANZ, A. (1958), . Amer. med. Ass., 167, 814.GOODKIND, M. J., HARVEY, R. M., and RICHARDSON,

D. W. (1958), Amer. J. med. Sd., 235, 164.HANLEY, T., and PLATTS, M. M. (g956), Lancet, i, 357.JOSE, A. D., and WOOD, P. (x958), Brit. med. y., i, 9.KATTUS, A., ARRINGTON, T. M., and NEWMAN, E. V.

(1952), Amer. J. Med., 12, 319.KAWAGA, C. M., CELLA, J. A., and VAN ARMAN, C. G.

(1957), Science, 126, 115.KERR, D. N. S., READ, A. E., HASLAM, R. M., and

SHERLOCK, S. (x958), Lancet, ii, 1084.LARAGH, J. H., HEINEMANN H. O., and DEMARTINI, F. E.

(I958), J. Amer. med. Ass., 166, I45.LEUTSCHER, J. A., NEHER, R., and WETTSTEIN, A. (I954),Experientia (Basle), 10, 456.LEUTSCHER, J. A., and JOHNSON, B. B. (x944), J. din. Invest.,

33, 1441.LIDDLE, G. W. (I957), Science, x26, Ioi6.MACHAFFIE, R. A., HUMOLLER, F. L., and WALSH, J. R.

(1958), Amer.C. rdd. , 505.MACKIE,J. E. STORMONT, J. M. HOLLISTER, R. M., andDAVIDSON, C. S. (I958), NewEngl. y. Med., 259,I151.MARTIN, J. D., and MILLS, I. H. (z956), Brit. med; ., i, 571.MATHESON, N. A., and MORGAN, T. N. (x958), Lanet, , 1195.McALLEN, P. M. (1955), Brit. Heart ., 17, 5.MOYER, J. H. FORD, R V., and SPURR, C. L. (I957), Proc. Se.exp. Biol. (N.Y.), 95, 29.PATEK, A J., MANKIN, H., COLCHER, H., LOWELL, A.,and EARLE, D. P. (z948), .. din. Invest., 27, 135.PEARSON, J. R., BINDER, C. I., and NEBER, J. (z955),J. Amer.med. Ass., I57, 339.PERKINS, J. G .PETERSON, A- B., and RILEY, J. A. (I95o),A-C..Y. Med., 8, XIS.PERSKY, L., CHAMBERS, D., and POTTS, A. (1956), .. Amer.med. Ass., z6I, 1625.PITTS, R. F., and SARTORIUS, O. W. (I950), Pharmacol. Rev.,2, z6x.PLATTS, M. M., and HANLEY, T. (x956), Brit. med. .., i, o078.PREEDY, J. R. K., and RUSSELL, D. S. (I953), Lancet, ii, itSi.RADO, J., and BLUMENFELD, G. (I958), Acta. med. scand.,x60, is.READ, A. E., HASLAM, R. M., LAIDLAW, J., and SHERLOCK,s. (x9s8), Brit. med. .,Y, 963.REISNER, E. H. Jr., and MORGAN, M. C. (I956), J. Amer. med.Ass., z60, 206.RELMAN, A. S LEAF, A., and SCHWARTZ, W. B. (x194),New Engl.J. Med., 250, 80o.ROSENHEIM, M. L., and SPENCER, A. G. (I956), Lancet, ii, 313.

ROSS, I. P., and ROSS, J. H. (1957), Ibid., i, 354.SALASSA, R. M., MATTOX, V. R., and POWER, M. H. (x958),Y. dcn. Endocr. Metab., 18, 787.SCHREINER, G. E., and BLOOMER, H. A. (1957), Ntew Engl J.Med., 257, Ioi6.SETTLE, E. (1957), Postgrad. med. J., a2, I86.SILVERMAN, J. J., and WORTHEN, J. F. (1952), 7. Amer. med.

Ass., 48, 200.SLATER, J. D. H., and NABARRO, J. D. N. (1958), Lancet, , 2z4.SMART, G. A. (1958), Postgrad. med. Y., 34, 325.SPENCER, A. G., and LLOYD-THOMAS, H. G. L. (i953),Brit. med. ., i, 957.SPENCER, A. G., and LLOYD-THOMAS, H. G. L. (x954)bidW, i, 597.STOCK, R J., MUDGE, G. H., and NURNBERG, M. J. ('95t),Circulbati (N. Y.), 4, S4.THORN, G. W., ARMSTRONG, S. H., DAVENPORT, V. D.,WOODRUFF, L. M., and TYLER, P. H. (1945), Y. din.

Invest., 24, 802.WALKER, A. M., SCHMIDT, C. F., ELSOM, K. A., and

JOHNSTON, C. G. (I937), Amer. . Physiol., x8, 95.WATSON, W. C., THOMPSON, T. J., and BUCHANAN, J. M.(I958), Lancet, 4 IXg.WEBSTER, L. T., and DAVIDSON, C. S. (1956), Proc. Soc. exp.Biol. (N. Y.), 91, 27.WOLFF, H. P., and KOCZOREK, Kh. R. (I955), Klin. Wschr.,

33, 1104.WOLFF, S. J., FAST, B., STORMONT, J. M., and DAVIDSONC. S. (1957), New Engl. J. Med., a57, 215.YATES-BELL, J. G. (x958), Brit. med. y., ii, 1392.

AddendumFurther investigations of the properties of

hydrochlorothiazide (6-chloro-7-sulphamyl - 3, 4 -

dihydro- I, 2, 4- benzothiadiazine- i, I - dioxide)have confirmed that the hydrogenated derivativeof chlorothiazide is usually about ten times morepotent than the parent substance, although thereare individual variations, in promoting a diuresisin patients with chronic heart failure (Fleminget al., 1959) or in those with ascites due to liverdisease (Kerr et al., I959). Since the action ofhydrochlorothiazide is more prolonged than thatof chlorothiazide in the majority of cases, a singledose in any one day should be adequate, whilechlorothiazide usually needs to be given twice inany one day (Platts, I959). Although increasedpotency based merely on a weight-for-weightbasis is of no particular merit in itself, it may bethat the very rare toxic effects on the bone marrowthat have been reported with chlorothiazide(Nordquist et al., 1959) will be even less likely tooccur when a smaller total dose of a compoundof similar chemical structure is used. Fromstudies on the patterns of diuresis it appears thatwith the exception of rather less bicarbonate ex-cretion after hydrochlorothiazide, the urinaryelectrolyte pattern produced by both drugs is thesame. Potassium loss may occur just as readilywith hydrochlorothiazide as with chlorothiazide(Havard and Fenton, 1959). Hypokalaemia ismore likely to occur in patients previously sub-jected to a low dietary sodium intake. In patientswith oedema of recent onset and not treated bysodium restriction the urinary potassium excretionmay not be materially increased by either drug.The maximum effective dose of hydrochloro-

Protected by copyright.

on March 5, 2020 by guest.

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.35.409.631 on 1 N

ovember 1959. D

ownloaded from