701

reversal was naturally slower but was achieved in 8-20weeks in all of 14 cases.The uninterrupted downward course of the quantita-

tive Wassermann reactions in all the cases treated bythis method is highly encouraging, contrasting with theupward or irregular direction eventually assumed in othercases recorded in the table, unsuccessfully treated byother regimens. Judged by the trend of the quantitativeWassermann reactions, this system of penicillin adminis-tration might be tentatively regarded as more hopefulthan the original scheme of 30,000 units 3-hourly for 10days, since after the latter treatment in 2 cases out of 3(the 4th becoming reinfected at an early date) the reactiontemporarily became more strongly positive, or returnedto positive, before finally settling down to a sustainednegative.At the present stage we certainly do not advocatethis precise form of penicillin treatment for general orroutine adoption, but the results, together with thosereported by Lloyd-Jones and Maitland (1945), do at leastshow that some such ambulatory scheme, involving avery few injections daily-perhaps even a single injectiona day-will prove to be effective in the large-scale treat-ment of early syphilis in civilian practice. † Lloyd-Jones

and Maitland gave a single intravenous or intramuscularinjection of 300,000 or 500,000 units a day to a total of2,400,000-5,000,000 units. Of 65 cases treated by theintravenous method (40 followed for 3-8 months, and 25for less than 3 months) there were 4 relapses, 7, 9, 9, and15 weeks after the end of treatment. It seems likely thata higher dose per injection, or a higher aggregate dosage, ’will give better results.

SUMMARY

It is again emphasised that penicillin treatment ofearly syphilis will not be appropriate for routinecivilian practice until suitable ambulatory courses areestablished.Reasons, supported by laboratory experiment, are

given for the tentative choice of an ambulatory courseof 600,000 units intramuscularly every hour for 3 doseson 5 successive days, although authoritative opinions areopposed to such a high dose at each injection.Our earlier work on penicillin in Spirillum minus

infections of mice demonstrated the curative value of alarge dose per injection, as against the merely temporaryeffect of a lower dosage.Under the conditions of our in-vitro experiments with

Spirochœta recurrentis, more than 99% of. the organismsare destroyed at 37° C by contact with 2-5 units of peni-cillin per ml. for 2-4 hours ; at higher concentrations thelethal action is more rapid. Three intramuscular injec-tions of 600,000 units at hourly intervals give rise toserum-penicillin levels of 5-20 units per ml. for 3-4hours.Although penetration of penicillin into the CSF cannot

be shown (in the absence of meningitis) after intramuscu-lar injection of the doses commonly used in standardschemes of treatment, some penetration is demonstrableafter 3 doses of 600,000 units at hourly intervals. Thissupports the thesis that such large doses are necessaryto ensure that penicillin reaches tissue spaces into whichit does not freely penetrate, and where spirochætes maybe sheltering.Preliminary results of treatment with 3 doses of 600,000

units at hourly intervals, given on 5 successive days, areexcellent. In all of 17 cases of primary and secondaryinfection symptoms and signs rapidly disappeared andthe blood-Wassermann reaction quickly became negative.It is premature to recommend adoption of this regimenfor routine’use, but some such ambulatory scheme mayconfidently be expected to prove suitable for the large-scale treatment of syphilis in civilian practice.Invaluable assistance has been provided by the Penicillin

Clinical Trials Committee (secretary, Prof. R. V. Christie)of the Medical Research Council, through whom supplies ofpenicillin were received; Dr. F. Glyn-Hughes, MO in charge VDwards ; Dr. C. McGibbon, assistant VD officer ; Mr. R. Ousby,SRN ; and Sister Fryer, Belmont Road Emergency Hospital.

t The cautionary note should be stressed, since a serological relapsehas occurred in one of the 14 secondary cases, while this paperhas been in the press.

References at foot of next column

DR. LOURIE AND OTHERS : REFERENCES

Beyer, K. H., Woodward, R., Peters, L., Verwey, W. F., Mattis,P. A. (1944) Science, 100, 107.

Bigger, J. W. (1944) Lancet, ii, 400.Chow, B. F., McKee, C. M. (1945) Science, 101, 67.Eagle, H., Musselman, A. D. (1944) J. exp. Med. 80, 493.Fleming, A. (1943) Lancet, ii, 434.Fleming, A., Young, M. Y., Suchet, J., Rowe, A. J. E. (1944) Ibid,

ii, 621.Florey, M. E., Florey, H. W. (1943) Ibid, i, 387.Garrod, L. P. (1945) Brit. med. J. i, 107.Lloyd-Jones, T. R. L., Maitland, F. G. (1945) Brit. J. vener. Dis.

21, 65.Lourie, E. M., Collier, H. O. J. (1943) Ann. trop. Med. Parasit. 37,

200. McDermott, W., Nelson, R. A. (1945) Amer. J. Syph. Gon. vener.

Dis. 29, 403. Moore, J. E., Mahoney, J. F., Schwartz, W. H., Sternberg, T. H.,

Wood, W. B. (1944) J. Amer. med. Ass. 126, 67.Rammelkamp, C. H., Bradley, S. E. (1943) Proc. Soc. exp. Biol.,

N.Y. 53, 30. Romansky, M. J., Rittman, G. E. (1944) Science, 100, 196.Ross, A. O. F., Nelson, R. B., Lourie, E. M., Collier, H. O. J. (1944)

Lancet, ii, 845.

EFFECT OF LARGE DOSES OF ALKALION KIDNEY FUNCTION

ARMY MALARIA RESEARCH UNIT,* OXFORD

THE investigations described here were undertaken tostudy in normal subjects the effects of the intensivealkali treatment that is still being recommended inblackwater fever, incompatible blood-transfusion, andcrush syndrome. For instance, the Medical ResearchCouncil (1944) recommended that in the treatment ofwound shock " all cases should be given sodium bicarbon-ate 7 g. (2 oz.) " (sic) " hourly by mouth until the urineturns red litmus blue, but not for longer than 24 hours "(italics are ours).

Intensive alkali therapy of this kind is accompanied bydangers, and its efficacy is open to question (Maegraithand Havard 1944). The results of the experimentsdescribed in this paper show that alkali, administered for24 hours to normal subjects at less than half the raterecommended above, produces sodium and waterretention and disturbances of renal function, whichbecome ’well-marked if the alkali is continued for 72hours.

EXPERIMENTAL

Alkali was given to three normal men in the following-doses : sodium bicarbonate gr. 60, sodium citrate gr. 60,aq ad. 1 oz. Sig. 1 oz. 2-hourly. Two subjects remainedon this course for 3 days, the other for one day only.

, The course was continued at night, so that 12 doses weregiven in the 24 hours. A fourth subject was given asimilar course containing an equivalent amount of sodiumin the form of sodium chloride. ’

_

Observations were made on the renal function, theacid-base balance, the electrolytes of the serum and urine,and the degree of hsemodilution and water retention.Control observations were made on each subject immedi-ately before the experimental period and sometimesafterwards also.

RESULTS ,

The results are shown separately for each subject intables I—IV. More complete protocols may be obtainedif required from B.G.M., Liverpool School of TropicalMedicine. Each day is reckoned from 10 AM. TheCO2 dissociation curve for whole blood from subject 1 isshown in the accompanying figure.

DISCUSSION .

Dosage.-The dosage given to the subjects of theseexperiments was high, but doses of this magnitude havebeen given in blackwater fever (Smith and Evans 1943) ;and, as pointed out above, much larger doses of sodiumbicarbonate have recently been recommended in crushinjuries.Renal Function.-The glomerular filtration-rate was

measured by the inulin clearance, and the effective plasma-flow by the diodone clearances (Smith et al. 1938, Alpert1941, McCance et al. 1944, Cole 1944). The effect of this course of alkali was to increase temporarily the glomeru-* Lieut.-Colonel B. G. Maegraith; Majors G. M. Brown, K. N. Irvine,

and R. J. Rossiter; Captains J. C. Lees, D. S. Parsons,C. N. Partington, and J. L. Rennie; Surgeon LieutenantR. E. Havard.

702

TABLE I-EFFECT ON SUBJECT 1 OF SODIUM BICARBONATE GR. 60 AND SODIUM CITRATE GR. 60 GIVEN.

TWO-HOURLY FOR 68 HOURS

. Administration of alkali commenced at 12 noon, June 26.

lar filtration-rate without any corresponding increase inthe effective plasma-flow, and then to produce a progres-sive impairment of tubular function.Thus in subject 3 after only 24 hours on the regime

there was an increase of 56% in the inulin clearance, andthe urea clearance was increased to the same extent (61%),probably owing to the increased glomerular filtration-rate. In subject 1, however, after 68 hours on alkalithere was an increase of inulin clearance amounting to120%, but there was no corresponding change in theurea clearance. Since there was no change in diodoneclearance in either subject, the filtration fraction (inulinclearance/diodone clearance x 100) was therefore in-creased in both. Changes in the filtration fraction areoften observed in conditions of impaired kidney function(Black et al. 1941, Corcoran and Page 1944, Lauson etal. 1944). The administration of alkali caused a haemo-dilution of about 10%, but this is almost certainlyinsufficient to explain the whole of the observed increasein the inulin clearances (Peters 1935).

In subject 2 after 76 hours on alkali there was a generalimpairment of kidney function. The diodone clearancewas slightly lower than in the control period, and theinulin clearance was not raised, but the urea clearance

was.only about a quarter of its original value. The blood.urea (Conway and Byrne 1933), which had remainedconstant for the first 24 hours, had risen at the end of 76hours to three times its control value. It was nearlytwice its control value 3 days later and was still slightlyraised a fortnight later. A similar slow recovery ofraised blood-urea is seen in blackwater fever and in therenal failure of many other conditions (Maegraith 1944).The urea clearance in subjects 1 and 2 does not run

parallel with the inulin clearance as it does-in subject 3.Thus the increased urea clearance observed in subject 3,after 24 hours on alkali, becomes well marked in subject 1after 68 hours and completely reversed in subject 2 after76 hours on alkali. This relative failure to concentratethe urea of the glomerular filtrate may be explained by aprogressive impairment of tubular function which allowsmore urea to diffuse back passively into the blood-stream,a view that is supported by the increased blood-urea insubject 2.The increased glomerular filtration-rate following this

high dosage of alkali may be explained by alterations inrenal haemodynamics, so that the blood-flow favours theglomeruli at the expense of the tubules. The probabilitythat such redistribution may take place has been indi-

TABLE II-EFFECT ON SUBJECT 2 OF SODIUM BICARBONATE GR. 60 AND SODIUM CITRATE GR. 60 GIVENTWO-HOURLY FOR 76 HOURS

* Administration of alkali commenced at 8 AM, August 9.t The serum figures on August 9 were obtained from blood taken immediately before the experiment commenced.

703

TABLE III-EFFECT ON SUBJECT 3 OF SODIUM BICARBON-ATE GR. 60 AND SODIUM CITRATE GR. 60 TWO-HOURLYFOR 24 HOURS

The administration of alkali commenced at 10 AM, August 5.

cated by Corcoran and Page (1943). In this connexion itshould be remembered that the high oxygen consumptionof the epithelium of the tubules renders it speciallyliable to damage from anoxia (Fishberg 1939, Scarff andKeele 1943). - When large amounts of alkali are beingexcreted, the oxygen consumption would be especiallyhigh in that part of the tubule actively engaged in theregulation of the reaction of the urine-i.e., in the distalconvoluted tubules (Ellinger 1940). In a state of relativeanoxia arising from the concurrence of excessive workand diminished tubular blood-flow the damage wouldtherefore be likely to appear first in this site. This wouldprovide a possible explanation for the observation byMcLetchie (1943) that in the alkalosis accompanyinggastric tetany the main damage is in the distal convolutedtubules. It may be significant that similar damage

TABLE IV—EFFECT ON SUBJECT 4 OF SODIUM CHLORIDEGR. 80 TWO-HOURLY FOR 24 HOURS

*Admistration of sodium chloride commenced at 10 AM, Nov. 23. h

has been reported in the renal anoxia accompanyingpernicious anaemia (Fishberg 1939). -

These considerations suggest therefore that the damageto the epithelium of the tubules in subject 2, indicatedby the diminished urea clearance and by nitrogen reten-tion, may have been due in part to the state of relativeanoxia, brought about by a diminution of its normalblood-supply.

Acid-base Balance.-The effect of alkali on the acid-basebalance developed rapidly. The increase in the totalserum CO2 content (Conway 1939) was apparent at theend of 4 hours and had nearly reached its full value after 8hours (table 11). The return to normal was also rapid andwas nearly complete in 24 hours (table I). This increasein the total serum CO2 was closely reflected in theincrease observed in the total CO2 content of whole blood(tables i, 11, and III).In subject 1 the alveolar CO2 showed an increase of

22%, which was constant throughout the course of alkali.If the usual assumption is made that at rest the alveolarCO2 is in equilibrium with the free CO2 ’of the serum, thenthis increase of 22%, taken in conjunction with theincrease of 40% observed in the total CO2 content of theserum, can be shown by Henderson-Hasselbach’s

Carbon-dioxide dissociation curves on whole blood : subject 1.

equation to correspond to an increase of 0-06 in the pH ofthe serum. A change of this order was confirmed bydirect measurement with the glass electrode (Stadie et al.1931. Meakins and Davies 1920). In the other 3 subjectsit was impossible to obtain reliable - sarmples of alveolarCO;:, but the pH of the blood was measured directly withthe glass electrode. , Subject 3 showed similar changes inpH and CO2 content of the serum. Subject 2 showedslightly smaller increases in the CO2 content of theserum and whole blood and no significant change in thepH, although the renal function was seriously upset.

It has thus been shown that this course of alkali mayproduce, within 24 hours, an increase of up to 40% in thetotal CO2 content of the serum, accompanied by anincrease of 0.04-0.06 in the pH of the blood and, in the onesubject examined, by,an increase of 22% in the alveolarCO2. It therefore produces a fairly intense alkalosiswhich is not entirely compensated.

Hœmodilution and Water Retention.—In subject 1 thepossibility of heemodilution and water retention wassuggested by a fall-in the haematocrit value from 43% to32%. The weight, which was 162 lb. at the end of thecourse of alkali, had dropped to 157 lb. 24 hours later.In view of these observations, and of their importance inconnexion with the increase of the glomerular filtration-rate, it was decided to investigate the question ofhaemodilution more closely. In the other 3 subjectstherefore the intake of water was regulated and thehaemoglobin concentration, the red-cell count, the oxygen-

704

combining power of the blood, and the albumin andglobulin concentrations of the plasma were observed,and more frequent observations were made of the weight.

In subject 2 it was found at the end of the course ofalkali that the hsematocrit fell to 91% of its control value,the haemoglobin to 95%,† the red-cell count to 91%, theoxygen-combining power to 90%, and the plasma proteinto 93%. It was concluded that there had been a dilutionof the blood by about 10% of its original volume, withoutmuch diminution in the mean corpuscular volume. There was a corresponding increase of 2 lb. in weightduring the course of alkali and a loss of 5 Ib. when itterminated.

In subject 3 there appears to have been a similar dilu-tion of the blood by about 10% of its previous volume.This subject gained 4 lb. in weight and lost it subse-quently.

In subject 4, who was given sodium chloride, there wasa hsemodilution of about 5% of the original volume, butthe mean corpuscular volume had diminished to about90% of the original figure, indicating some shrinkage ofred cells. There was no significant change of weightduring the experiment.The course of alkali thus increased the blood volume bv

about 10%, which corresponds to a retention of about 700c.cm. of water for blood alone and of some 2000 c.cm. iftissue fluids are also taken into account. This is in goodagreement with the amount indicated by the increase inweight-i.e., about 4 lb.

During the course of alkali the daily sodium intakewas increased by 775 mille-equivalents over the normalintake of about 125 mille-equivalents, making a totaldaily intake of about 900 mille-equivalents of sodium.The figures of sodium excretion during the course ofalkali therefore suggest considerable sodium retention.In subject 1 this retention amounted to about 1000mille-equivalents of sodium at the end of the 3 days. Ifthe extracellular fluid is estimated as 20 litres, 200mille-equivalents would be required to raise the sodiumconcentration in the plasma and tissue fluids from 140 to150 mille-equivalents per litre (table I). The remaining800 mille-equivalents would require a water retention ofabout 5 litres for a final concentration of 150 mille-equivalents per litre to be attained. A similar calcula-tion for subject 2 indicates a water retention of about 6litres, for subject 3 a water retention of about 1 litre, andfor subject 4 a water retention of about 0-6 litres. Thedegree of water retention calculated by this roughmeasure is of the same order as that indicated by thechanges in weight and the dilution of body fluids.Serum Electrolytes.-All subjects showed an initial

increase in sodium concentration (McCance and Shipp1932), which was followed by a return towards normal insubjects 1 and 2 as the course of alkali proceeded. Thisreturn to normal concentration was probably due tothe progressive haemodilution. With the exception ofsubject 4 the potassium concentration (Hoffman 1937)fell, though in subject 2 it rose again towards the end ofthe experiment. This fall may be either an adjustmentto the increase in sodium or due simply to the increasedloss of potassium in the urine. It may be significantthat in subject 4, who had no increased excretion ofpotassium, there was no fall in the serum potassium.In this subject the chloride (Sendroy 1937) and sodiumboth showed the degree of increase that would be expectedfrom the dosage.

In general the change in serum electrolytes in allsubjects followed the expected course (Peters and VanSlyke 1931, Baird and Haldane 1922).

General Clinical and Subjective Conditions.—There wereno obvious subjective effects from this course of alkaliuntil it had been taken for 2 or 3 days, but on the 3rdday its effects were striking. They were restlessness.headache, anorexia, and some evidence of deterioration inmental powers. No significant changes in blood-pressurewere observed. Both subjects 1 and 2 looked and feltill on the last day of the experiment, but neither subjects3 and 4 experienced any severe symptoms of this nature.

. GENERAL CONCLUSIONS

It is evident that this course of alkali (sodium bicar-bonate gr. 60 and sodium citrate gr. 60 2-hourly) is

t These figures all refer to % of control values.

excessive and probably would be dangerous if continuedbeyond 3 days, even in a healthy adult. It quicklyproduces fairly intense alkalosis that is not entirelycompensated. There is an increase in glomerular filtra-tion at the end of 24 hours which can be explained by aredistribution of blood-flow. This increased filtration-rate may be beneficial in the early stages of a course ofalkali, but by the end of 24 hours there is already someretention of sodium. By the end of 3 days the disturb-ance of kidney function may be sufficiently profound toproduce some degree of renal failure, indicated by adecreased urea clearance and an increased blood-urea.There is- considerable retention of sodium, the kidneyapparently being unable to excrete sodium at the raterequired to balance the intake. This leads to retention ofwater and considerable dilution of the blood and tissuefluids.

If these effects can be induced in healthy adults by ashort vigorous course of alkali, it is clear that in black-water fever and other conditions often accompanied byrenal failure, alkalis, if given at all, should not be pushedto extremes.

SUMMARY

Three subjects were given an intensive course of alkaliconsisting of sodium bicarbonate gr. 60 and sodiumcitrate gr. 60 2-hourly. One subject received this coursefor 1 day ; two subjects for 3 days. -

All the subjects on alkali had disturbances of renalfunction. In one, on the course for 3 days, this was wellmarked and led to a three-fold increase in blood-urea.

All the subjects on alkali developed fairly intensealkalosis, not entirely compensated, together with sodiumand water retention. There was dilution of the blood byabout 10%.Both the subjects on alkali for 3 days complained of

symptoms.A fourth subject was given an equivalent amount of

sodium as sodium chloride for 24 hours. He showed aslight degree of sodium and water retention and noalteration of renal function.The amount of alkali given in these experiments had

measurable physiological effects, including disturbancesof renal function. We consider, therefore, that largedoses of alkali should not be administered in conditionswhere renal failure may supervene.We express our thanks to Brigadier F. A. E. Crew, FRS,

director of biological research, for permission to publish thiswork ; Prof. C. C. Douglas, FRS, for his advice in the design ofthe experiment, the loan of apparatus, and for performingsome of the analyses ; and Prof. R. A. McCance and Dr. S. W.Cole for their help and advice.

REFERENCES

Alpert, L. K. (1941) Bull. Johns Hopk. Hosp. 68, 522.Baird, M. M., Haldane, J. B. S. (1922) J. Physiol. 56, 259.Black, D. A. K., Powel, J. F., Smith, A. F. (1941) Ibid, 99, 344.Cole, S. W. (1944) personal communication.Conway, E. J. (1939) Microdiffusion Analysis and Volumetric Error,

London, p. 176.— Byrne, A. (1933) Biochem. J. 27, 419.

Corcoran, A. C., Page, I. H. (1943) J. exp. Med. 78, 205.— — (1944) Ann. intern. Med. 21, 747.

Ellinger, P. (1940) Quart. J. exp. Physiol. 30, 205.Fishberg, A. M. (1939) Hypertension and Nephritis, Philadelphia.Hoffman, W. S. (1937) J. biol. Chem. 120, 57. Lauson, H. D., Bradley, S. E., Cournard, A. (1944) J. clin. Invest. 23,

387.McCance, R. A., Shipp, H. L. (1932) Biochem. J. 25, 1845.

— Young, W. F., Black, D. A. K. (1944) J. Physiol. 102, 415.McLetchie, N. G. B. (1943) J. Path. Bact. 55, 17.Maegraith, B. G. (1944) Trans. R. Soc. trop. Med. Hyg. 38, 1.

— , Havard, R. E. (1944) Lancet, ii, 338.Meakins, J., Davies, H. W. (1920) J. Path. Bact. 23, 451.Medical Research Council (1944) War Memorandum No. 1, 2nd ed.Peters, J. P. (1935) Body Water : The Exchange of Fluids in Man,

Springfield.— Van Slyke, D. D. (1931) Quantitative Clinical Chemistry,

Baltimore.Scarff, R. W., Keele, C. A. (1943) Brit. J. exp. Path. 24, 147.Sendroy, J. (1937) J. biol. Chem. 120, 405.Smith, F., Evans, R. W. (1943) Brit. med. J. i, 279.Smith, H. W., Goldring, W., Chasis, H. (1938) J. clin. Invest. 17, 263.Stadie, W. C., O’Brien, H., Laug, E. P. (1931) J. biol. Chem. 91, 243.

ON Thursday, Dec. 6, at 2.30 PM, a demonstratioa. Qf thetechnique of the use of a variety of contraceptive methods willbe given at the CBC Mothers’ Clinic. Medical practitionersand senior students who wish to attend should apply in writingto the hon, secretary, 108, Whitfield Street, London, Wl.