J. clin. Path., 1971, 24, 419-429

Use and control of cephalosporinsSUSANNAH EYKYN

From the Department of Clinical Microbiology, St Thomas's Hospital, London

It is now over 20 years since Professor Brotzurecovered the mould Cephalosporium acremoniumfrom Sardinian sewage and found that it producedantibiotic material. Figure 1 represents a simplifiedcephalosporin family tree. The first antibiotic to beisolated from the mould was cephalosporin N, anacyl derivative of 6-aminopenicillanic acid, sus-ceptible to the action of stapylococcal penicillinaseand much less active against Gram-positive bacteriathan benzyl penicillin. Cephalosporin P consists ofone major component and at least four minorcomponents: its major component is a steroidrelated chemically to, but considerably less activethan, fusidic acid. The most important antibioticelaborated by the cephalosporium mould iscephalosporin C, but it is only produced in minutequantities. It is from cephalosporin C that the

semi-synthetic cephalosporins have been derived.Several hundred analogues have been produced bysubstitution of different chemical groups at the 7 and3 position (R1 and R2 in the diagram) of the cepha-losporin nucleus and have been investigated forantimicrobial activity. Four cephalosporins havebeen extensively studied in various parts of theworld-cephaloridine, cephalothin, cephalexin, andcephaloglycine-and of these the first three drugsare now available commercially in Britain and it iswith them that this paper is concerned. Cephaloridinehas been available since 1964 and cephalothin andcephalexin much more recently.

Antibacterial Spectrum

The antibacterial spectrum of the three cephalo-

CEPHALOSPORIUM ACREMONIUM

cephalosporin N cophalosporin P

CEPHALOSPORIN C

CEPHALOSPORIN NUCLEUS

SR,HN

CO2H

-.T -1 -ICEPHALORIDINE CEPHALOTHIN CEPHALEXINFig. 1 Derivatives of Cephalosporium acremonium.

419

CEPHALOGLYCINE

420

sporins is similar but they differ markedly in theiractivity. In general, cephaloridine is the most activeand cephalexin the least active though unexpecteddifferences may be found for some organisms.Cephalosporins exhibit a wide range of activityagainst Gram-positive and Gram-negative bacteriabut are ineffective against Pseudomonas aeruginosa,Enterobacter, indole-positive Proteus and Bacter-oides, and of limited activity against Haemophilusinfluenzae and Streptococcus faecalis. A large numberof minimum inhibitory concentrations (MICs) havebeen determined for the three cephalosporins inparallel on strains recently isolated at St. Thomas'sHospital. Both heavy and light inocula of theorganisms were used on solid media: a heavyinoculum consisted of approximately 0-02 ml of abroth culture containing 108 organisms per ml anda light inoculum consisted of a 10-' dilution of theabove.

COAGULASE-POSITIVE STAPHYLOCOCCIThe results obtained for 53 coagulase-positivestaphylococci are shown in Figure 2. Cephaloridineis the most active cephalosporin against penicillin-sensitive staphylococci: its activity compares favour-ably with that of penicillin. Cephalothin is onlymarginally less active than cephaloridine, withcephalexin least active. The majority of strains haveMICs of cephalexin about 30 times those of

Susannah Eykyn

cephaloridine. With all three cephalosporins there isvery little difference between the MIC value obtainedwith heavy and light inocula for penicillin-sensitivestaphylococci.However, the results are rather different for the

penicillinase-producing strains: some of the strainsinvestigated were resistant to streptomycin, tetra-cycline, and erythromycin as well as to penicillin.With cephaloridine, there is a marked inoculumeffect: the MICs obtained with a light inoculum arevery much less than those with a heavy inoculum.The inoculum effect is much less evident with cepha-lothin and cephalexin. Against a heavy inoculum ofpenicillinase-producing staphylococci, cephalothinis the most active cephalosporin due to its greaterstability to staphylococcal penicillinase. Cephalexinis also more stable to staphylococcal penicillinasethan cephaloridine but is intrinsically less activethan cephalothin.

Twenty-three strains of methicillin-resistantstaphylococci have also been investigated forcephalosporin sensitivity. With a light inoculum themajority of these strains were inhibited by 1V5 ,ugcephaloridine per ml and by 0-78 ug cephalothinper ml, but when a heavy inoculum of the organismwas used, all strains were highly resistant tocephaloridine but much less resistant to cephalothin.All strains tested were highly resistant to cephalexinboth with heavy and light inocula. Some degree of

pa antibiotic per mlinocukw 25 12-5 625 34 t58 078 039 0-19 0-09 004 002 total

heavy _ 17cephaloridine -.g._1

light ~ m 17l-

heavy 17z

w cephalexhin 0 0 .*...,....... |1light em17

-_ hory 17.17ccephalexin

light em ~~~~~~~~~~~~~17

hievyC0i_ 36cephalorkldne.ig*5g*5g5

* seem 36

heay"gw~ e 36oscphslothin

light 36

Fig. 2 Effect ofinoculum size oncephalosporin MICfor coagulasepositive staphylo-cocci.

Use and control of cephalosporins

cross-resistance occurs between methicillin and allthree cephalosporins; this seems to be most markedbetween cephalexin and methicillin and least evidentbetween cephalothin and methicillin. Kind, Kestle,Standiford, Freeman, and Kirby (1969a) havedemonstrated that cephalexin, like methicillin, canproduce resistant variants of staphylococci onrepeated exposure of sensitive strains to either drugin vitro. These variants are seen as small colonieswithin the zone of inhibition around the antibioticdisc. In contrast, they never found isolated colonieswithin the zones of inhibition around discs ofcephaloridines or cephalothin and colonies selectedfrom around the inhibition zones remained sensitiveto these antibiotics and methicillin even after serialtransfer.

GRAM-NEGATIVE BACILLI ANDSTREPTOCOCCUS FAECALISThe MIC results obtained using a light inoculum of100 strains of Escherichia coliwith the cephalosporinsare shown in Figure 3. Cephaloridine is the mostactive; over 60% of the strains were inhibited by 3 9,ug per ml or less whereas only 30% were inhibitedby the same amount of cephalothin or cephalexin.The results for 85 strains of Klebsiella (all non-motileand lysine decarboxylase positive) are shown inFigure 4 (light inoculum). There is little to choose

Fig. 3 Sensitivity of 100 strains of E.50- cephalosporins.

40

7.30 7 U2-.

U)'I

between the activity of the three cephalosporinsagainst Klebseillae, with cephalothin perhapsmarginally best. The indole-positive Proteus seemto be universally resistant to cephalosporins but themajority of Proteus mirabilis strains are sensitive,with cephaloridine and cephalothin about equallyactive and cephalexin least so. The results obtainedusing light inocula of 66 strains of Proteus mirabilisare shown in Figure 5.The MICs for 100 strains of Streptococcus faecalis

(see Fig. 6) were lower for cephaloridine than forcephalothin, but with values of 7-8 to 15-6 ,ug per mlfor the majority of strains using light inocula, neitherantibiotic could be considered very active. Moststrains were highly resistant to cephalexin and over50% of these required 125 ,ug per ml for inhibition.A decrease in inoculum size almost always led to

a decrease in cephalosporin MIC for the Gram-negative bacilli. The actual degree of decrease inMIC obtained was rather variable and was some-times as much as a hundredfold or even more. Forany individual organism, however, the inoculumeffect was not necessarily of the same magnitude foreach cephalosporin tested. It is perhaps associatedwith the type and amount of fl-lactamase producedby each organism. Figure 7 compares the effect ofinoculum size on the MIC of cephaloridine andcephalexin for 76 strains of Escherichia coli.

.coli toD CEPHALORIDINE

XII CEPHALOTHIN

E: CEPHiALEXIN

0-48

pg antibiotic per ml

421

Susannah Eykyn

60*

50

30-

20-

10-

III I

I

I048 0-97 1 95

Fig. 4

70-

.1 40

a.5._

20

I0

Fig. 5

HI

1I1 I097 195

3-9I ll Ii1 J Sr-i Fl [n.LNItr ,I

7-8 15-6 31- 2

pg antibiotic per ml

17-8

62-5 125 250 '>500

UaU

ll ii L3 15-6 1 31-5

n, U ,

CEPHALORIDINE

CEPHALOTHIN

CEPHALEXIN

62-5 X 125 1 250 so5 I

pg antibiotic per ml

U

a

CEPHALORIDINE

CEPHALOTHIN

CEPHALEXIN

Fig. 4 Sensitivity of85strains of Klebsiella tocephalosporins.

Fig. 5 Sensitivity ofj66strains ofProteus mirabilis tocephalosporins.

IN...

I !! lim A! lim li. li. R-I fi 0, - lis-!M. -7- -T ! !!. Z-!!LLI

., _ , ,_ _LI-ALi_ _ _ _ _ wI AELF1I

422

0

'P

IsI0--ol

1-

Use and control of cephalosporins 423

70-

60-

50-

40-

30-

20-

10-

0 0o-97r-

Fig. 6

UaEli

CEPHALORIDINECEPHALOTHINCEPHALEXIN

Fig. 6 Sensitivity of 100strains of Streptococcusfaecalis to cephalosporins.

Fig. 7 Effect of inoculumsize on MIC of cephalosporinsfor 76 strains of E. coli.

"~~~~~-M IIn "I1|1lm7'8 'Js 315 62-5 125 2s0250

pg antibiotic per ml

CEPHALORIDINE HEAVY INOCULUM

LIGhT INOCULUM

[] CEPHALEXIN HEAVY INOCULUM

LIGHT INOCULUM

0

._a

0(9

pg antibiotic per mlFig. 7

0

._i0s0

424

The effect of inoculum size on cephalosporinMIC is really only of therapeutic importance forpenicillinase-producing staphylococci. Since cepha-loridine is to some extent destroyed by penicillinaseit is probably not a suitable drug to treat severeinfection caused by staphylococci producing largeamounts of the enzyme. Although in theory ifenough cephaloridine was given it might be effective,cephalothin would be the better drug in this situation.

Pharmacological Properties (see Fig. 8)

MODE OF ACTIONAll cephalosporin derivatives act by inhibition of

Susannah Eykyn

bacterial cell wall synthesis and are bactericidal: ingeneral, the level of drug required to achieve abactericidal effect for most organisms is only two tothree times that required for bacteriostasis.

ROUTE OF ADMINISTRATIONCephaloridine and cephalothin are only very poorlyabsorbed from the gut and must be given by injectionwhich may be intramuscular or intravenous. Theserum levels achieved with cephaloridine are about50% higher than those reached with an equivalentdose of cephalothin. Cephalexin and cephaloglycineare absorbed from the gut and cephalexin in par-ticular is very well absorbed; it is available for oral

Cephaloridine Cephalothin Cephalexin

Mode of action Inhibition of Cell Wall SynthesisRoute of administration Intramusucular Intramuscular Oral

Intravenous IntravenousBinding to serum protein 20% 65% 13-19%Excretion Mostly by glomerular filtration; Mainly by tubular secretion Secreted and reabsorbed

minimal tubular secretion by renal tubuleProbenecid Minimal effect Blocks secretion Blocks secretionToxicity < Rashes

Dose-related nephrotoxicity, Pain on intramuscular injection. Diarrhoea, vulvovaginitispossibly enhanced by frusemide Thrombophlebitis on intravenous

injection. Positive direct Coombs.? ? nephrotoxic

Fig. 8 Comparison of cephalosporins.

E._

Q

0)8n

Fig. 9 Serum cephalexin levels in fasting volunteersfollowing a dose of 250 mg.

- -N.

3 4

time in hours

,t

Use and control of cephalosporins

administration only. In a recent study in our lab-oratory, 16 fasting volunteers (laboratory tech-nicians) were given an oral dose of 250 mg ofcephalexin and their serum levels were measuredover the next four hours; not only was the peakserum level obtained very variable in individuals(levels from 2-7 to 19-0 ,ug cephalexin per ml) but thetime taken to reach the peak level was also veryvariable. Four of the absorption curves obtained areshown in Figure 9. Since the volunteers all took theoral dose under identical conditions it is difficult toexplain these discrepancies but other workers havedescribed similar findings (Brumfitt, 1970). Moreoverthis variability of time and height of peak serumlevel occurred with both preparations of cephalexin(Keflex and Ceporex).

BINDING TO SERUM PROTEINCephalosporins do not all bind to serum proteins tothe same extent. Cephaloridine and cephalexin areonly minimally bound but a large proportion ofcephalothin is bound. Although the degree ofprotein binding found with cephalothin is muchgreater than that of ampicillin or methicillin,it is still considerably less than that of cloxacillin.The actual figures given in the literature for proteinbinding of cephalosporins are rather variable butKind, Kestle, Standiford, and Kirby (1969b) quotevalues of 20% for cephaloridine, 13-19% forcephalexin, and 65% for cephalothin.

EXCRETIONCephalosporins are almost entirely excreted by thekidney, but they are not all handled by the kidneyin the same way. Cephaloridine is excreted mainlyby glomerular filtration with only a minimal amountof tubular secretion; about 75% of a single dose ofcephaloridine can be recovered from the urine inabout eight hours (Weinstein and Kaplan, 1970).The effect of probenecid on cephaloridine serumlevels is negligible unless large doses of cephaloridineare given.

Cephalothin is excreted mainly by tubularsecretion and this can be effectively blocked byprobenecid. However, about 20% of the cephalothinappearing in the urine is present as the weaklyantibacterial metabolite desacetyl cephalothin dueto drug metabolism in the liver. Cephalothin isexcreted rather more rapidly than cephaloridine:over 95% of an administered dose can be recoveredfrom the urine within six hours (Weinstein andKaplan, 1970). Cephalexin is secreted and re-absorbed by the renal tubule and its excretion, likethat of cephalothin, can be blocked by probenecid.Excretion of cephalexin is rapid and 80-100% of an

425

administered dose can be recovered from the urinein six hours.

TOXICITYAll three cephalosporin derivatives are of relativelylow toxicity; they should, however, be administeredwith caution to patients with a history of penicillinhypersensitivity, since about 10% of such patientswill also prove hypersensitive to cephalosporins.

NephrotoxicityCephaloridine is known to be a potentially nephro-toxic drug both in laboratory animals and in man.In animals, the dose required to produce renaldamage varies widely with different species and thefemale rabbit is the most sensitive. Cephaloridinecauses necrosis of the proximal cells of the kidneytubules (Atkinson, Caisey, Currie, Middleton, Pratt,Sharpe, and Tomich, 1966a; Atkinson, Currie,Davis, Pratt, Sharpe, and Tomich, 1966b). Lawson,Macadam, Singh, Gavras, and Linton (1970) havedemonstrated that in the rabbit the nephrotoxiceffect of cephaloridine is enhanced by the diureticfrusemide and suggested that this may also be true inman. Foord (1969) reported 36 cases in whichcephaloridine was suspected of causing renal failurein man; in nine of these the renal function was knownto be impaired before administration of the drug andin 13 others it was probably impaired. In the eightpatients in Foord's series whose serum cephaloridinelevel was measured, figures from 40 to 1,500 ,ug perml were recorded. Seven patients also receivedfrusemide. The nephrotoxic effect of cephaloridineon the proximal renal tubule is probably dose-related and it is most likely to occur if serum levelsexceed 100 to 150 ,ug per ml. It is potentially re-versible on stopping the drug.

Figure 10 summarizes a case of probable cepha-loridine nephrotoxicity seen at St. Thomas'sHospital. The patient was a 54-year-old woman withdiabetes who was operated on for replacement ofthe mitral and aortic valves on 16 July 1970. Shereceived prophylactic cephaloridine in a dose of 500mg six hourly intravenously.Her renal function was known to be impaired at

the time of operation Lnd blood urea levels of 50 to85 mg per 100 ml had been recorded over theprevious year. She was treated postoperatively withintravenous frusemide on several occasions. Herurine output started to fall on the seventh post-operative day and at the same time the blood ureaconcentration rose to nearly 400 mg per 100 ml.On the 11th postoperative day she became shockedand was found to have Enterobacter aerogenes in herblood and the same organism was recovered,amongst others, from her sputum. The Enterobacter

Susaninah Eykyn

blood culturesEnterobacter aerogenes

operation

iEAM£IC-

IV CEPHALORIDINE 5 TRIMETHOPRIMn SULP/ ETHOXAZOLEV EHLORIDINE 5QO0"g ¶JL intravenous 7 LQ[-oraI

Fig. 10 Possible cephalori-dine nephrotoxicity in a 54-vear-old woman tollowingaortic and mitral a/lvereplacement.

t FRUSEMIDE

PERITONEAL DIALYSIS

mJULY AUGUST

was highly resistant to cephaloridine which was thendiscontinued. She was dialysed by the peritonealroute from the 12th to the 26th postoperative dayand her renal function recovered its preoperativelevel. The septicaemia was successfully treatedinitially with gentamicin and later with intravenousand then oral trimethoprim and sulphamethoxazole.It is possible that the complicating septicaemia mayhave contributed to the renal failure, but it is alsohighly likely that the combination of renal impair-ment, cephaloridine, and frusemide were at leastpartly to blame, since her urine output fell and herblood urea rose several days before she becameshocked. Unfortunately cephaloridine serum levelswere not estimated, but two days after stopping thedrug there was considerable residual cephaloridinein her serum as this interfered with the assay ofgentamlcin.

Cephalothin causes minimal swelling of theproximal tubular epithelium in rabbits whereas theequivalent dose of cephaloridine produces tubularnecrosis (Perkins, Apicella, Lee, Cuppage, andSaslaw, 1968). Lawson et al (1970) did not, how-ever, find that frusemide, given in conjunction withcephalothin to rabbits with kidneys minimallydamaged by glycerol, produced further nephro-

toxicity. Despite the encouraging reports fromanimal work there are now suggestions that cepha-lothin, like cephaloridine, may be nephrotoxic inman. In reported cases the dose of cephalothin hasusually been high but it has not always been easy toassess to what extent cephalothin alone was respon-sible for the deteriorating renal function as otherfactors were frequently involved (Rahal, Meyers, andWeinstein, 1968; Turcotte, Herrmann, Haig, O'Dell,Cerny, and Greene, 1970; Pickering, Spooner.de Quesada, and Cade, 1970). There are as yet noreports suggesting that cephalexin is nephrotoxic.

Positive direct Coombs testA positive direct Coombs test is found not in-frequently in patients being treated with cephalothin,particularly when the serum levels of the drug arevery high or when the patient has some degree ofrenal impairment or hypoalbuminaemia (Gralnick.Wright, and McGinniss, 1967; Molthan, Reidenberg,and Eichman, 1967). It is not associated withhaemolytic anaemia and seems to be of no clinicalsignificance buLt may interfere with cross-matchingof blood in the laboratory. A positive direct Coombstest has also been reported in patients on cepha-loridine (Fass Perkins, and Saslow. 1970) but the

400

15001

2E

E 100(200

,MO _

=10015001

426

Use and control of cephalosporins

8% incidence with this drug was much lower thanthe reported series for cephalothin where theincidence varied from 38 to 75 %. There was no

association with impaired renal function, hypo-albuminaemia, or high dosage therapy. Cephalexinmay also cause a positive direct Coombs, test though,as with cephaloridine, this is uncommon. Of 36patients treated with cephalexin for recurrent urinarytract infection at St. Thomas's Hospital, only one

developed a positive direct Coombs test duringtherapy.

Other toxic effectsCephalothin, unlike cephaloridine, is painful on

intramuscular injection particularly with a 1 g doseand may be intolerable by this route after severaldays and intravenous injection may cause thrombo-phlebitis. Cephalothin may also cause neutropenia.Cephalexin, like most oral broad-spectrum anti-biotics, may cause gastrointestinal disturbances butsome of the nausea and vomiting reported in earlierclinical trials may well have been attributable to theunpleasant smell and taste of the initial clinical trialmaterial. Vulvovaginitis has been reported by Leigh,Faiers, and Brumfitt (1970) and by Faireley (1970) as

a troublesome side effect of cephalexin.

CEPHALOSPORINS IN RENAL FAILURE

Since there exists a possible danger of nephrotoxicity,particularly with cephaloridine, it is important toknow how patients whose renal function is grosslyimpaired excrete cephalosporins. Table I shows theeffect of uraemia and of dialysis on the serum half-life of cephaloridine, cephalothin, and cephalexin.

In patients with normal renal function the serumhalf-life of cephaloridine and cephalexin is similar(1 2 hours and 1-5 hours) whereas cephalothin isexcreted more rapidly with a serum half-life of 0-5hours. In uraemia (that is in patients whose creatinineclearance is less than 3 ml per min), the serum half-life of cephaloridine and cephalexin is greatly

427

increased (23 hours and 215 hours). The serum

half-life of cephalothin in uraemia is prolonged(2-9 hours) but to a much lesser extent than cepha-loridine or cephalexin. It follows, therefore, that thedosage schedule for cephaloridine and cephalexinwill require greater modification in renal failurethan that for cephalothin. During dialysis the serum

half-life of cephaloridine and cephalexin is markedlyreduced strongly suggesting that they are both largelyremoved by dialysis. This is probably also true forcephalothin. There is also a tendency for thecephaloridine serum half-life to fall with increasingduration of maintenance haemodialysis (Curtis andMarshall, 1970). The reasons for this are not knownbut presumably extrarenal mechanisms of cepha-loridine degradation become increasingly importantwith increasing duration of maintenance haemo-dialysis. The interval between doses of the cepha-losporins will need to be much shorter when a patientis being dialysed than during the periods betweendialysis, or therapeutic levels of the drug may not bemaintained. It may well be necessary to monitorserum levels in uraemic patients who are treatedwith cephalosporins.

Assay of Cephalosporins

The overall methods are set out in Table II. Cepha-losporins can readily be assayed microbiologicallyby the plate method with reasonable accuracy.Cephaloridine is undoubtedly the easiest of the threedrugs to assay and gives large clear-cut zones ofinhibition on Oxoid DST agar using Bacillussubtilis ATCC 6633. Sarcina lutea can also be used,but in our experience gives less distinct zone margins.Stock solutions of cephaloridine can be stored at4°C for at least a week without loss of potency andpossibly longer.

Cephalothin can be assayed by the method de-scribed for cephaloridine but tends to give smallerzones of inhibition. It is important to remember also

Serum HalfLife Cephaloridine Cephalothin Cepkalexin

Normal renal function 1P5 ± 0 3 hre 05 + 0-3 hr 1-2 ± 0 4 hr'1-52 hr2 0 85 hr3

23 i 3 hr2 2-8 ± 0 9 hrs 21-5 i 1-9 hr2Uraemia: not on haemodialysis' 20 hr 3

10-4 hr4 2-9 hr3 21-8 hr'

Uraemia: on haemodialysis1 3-5 hr4 45 i 09 hr24-6 i 07 hrP

Table I Cephalosporins in renal failureLUraemia = Creatinine clearance < 3ml/min'From Kabins et al (1970)3From Kunin and Atuk (1966)4From Curtis and Marshall (1970)5From Bailey et al (1970)

Agar Assay Organism Buffer Stability in Solution

Cephaloridine DST B. subtilisIS. lutea pH 7 StableCephalothin DST B. subtilisiS. lutea pH 7 StableCophalexin Special" B. subtilisIS. lutea pH 6 Unstable

Table II Plate assay of cephalosporins5Oxoid agar No. 3 1-2%Oxoid peptone 0-5%Oxoid lab-lemco 0-3 %,Sodium citrate 1-0%/

that one is in fact assaying a mixture of cephalothinand its descetyl metabolite using a cephalothinstandard. The standard solutions for assay ofcephalothin in serum must be made up in humanserum as it is so highly bound to protein, but forcephaloridine and cephalexin comparable results areobtained whether serum or buffer is used to make upthe standard solutions. Stock solutions of cepha-lothin can be stored at 4°C for at least a week.

Cephalexin assay is unsatisfactory if DST agar isused: the best results are obtained with the 'specialagar' which is recommended by Glaxo Laboratories.Stock solutions of cephalexin seem to deterioraterapidly if stored for any length of time and it isalways advisable to make up fresh stock solutionsfor each assay.

Clnical Use of Cephalosporins

Cephaloridine is now well established as a valuablebactericidal parenteral antibiotic, with activityagainst a wide range of organisms. It is the mostactive of the available cephalosporin derivativesthough less stable to staphylococcal penicillinasethan cephalothin or cephalexin. The serum levelsachieved with cephaloridine are considerably higherthan those with cephalothin in equivalent dosage.Despite the possible risk of nephrotoxicity withcephaloridine it should not be withheld from apatient with impaired renal function if it istherapeutically indicated, but caution should beexercised in devising the dosage schedule and serumlevels of the drug monitored where necessary. Itshould not be given in conjunction with frusemide.

Cephalothin is not only less active than cepha-loridine, except against penicillinase-producingstaphylococci, but it also has the added disadvantageof producing pain on intramuscular injectionwhereas cephaloridine is relatively painless. Althoughit was claimed initially that cephalothin did notcause renal damage it now appears that this drug,like cephaloridine, may cause tubular necrosis. If aparenteral cephalosporin is clinically indicatedcephaloridine should be given in preference tocephalothin.

Cephalexin is the only oral cephalosporin availablein the UK: it is unfortunately much less active thancephaloridine and against most organisms is lessactive than cephalothin also. However, it is very wellabsorbed and causes relatively few side effects withto date no reports of nephrotoxicity.Numerous clinical trials have now been carried

out on cephalexin and it has been given for a widevariety of infections; it is, however, probably mostuseful for treatment of urinary infections (par-ticularly Klebsiella and Citrobacter) and for somepatients with pencillin hypersensitivity for whom anoral drug is indicated. Cephalexin is at presentextremely expensive and in most of the situationswhere it might be clinically indicated there is nearlyalways a much cheaper, equally effective alternativedrug available.

I would like to thank Eli Lilly and Glaxo for suppliesof cephalexin and for financial support for some ofthis work, Mr M. Braimbridge for permission tostudy his patient, Miss C. Jenkins, AIMLT, for veryefficient technical assistance, and Dr I. Phillips forhelpful advice and criticism.

References

Atkinson, R. M., Caisey, J. D., Currie, J. P., Middleton, T. R., Pratt,D. A. H., Sharpe, H. M., and Tomich, E. G. (1966a). Subacutetoxicity of cephaloridine to various species. Toxicol. appl.Pharmacol., 8, 407-428.

Atkinson, R. M., Currie, J. P., Davis, B., Pratt, D. A. H., Sharpe,H. M., and Tomich, E. G. (1966b). Acute toxicity of cepha-loridine, an antibiotic derived from cephalosporin C. Toxicol.appl. Pharmacol., 8, 398-406.

Bailey, R. R., Gower, P. E., and Dash, C. H. (1970). The effect ofimpairment of renal function and haeodialysis on serum andurine levels of cephalexin. Postgrad. med. J., October 1970,Supplement Vol. 46, 60-64.

Brumfitt, W. (1970). Personal communication.Curtis, J. R., and Marshall, M. J. (1970). Cephaloridine serum levels

in patients on maintenance haemodialysis. Brit. med. J., 1,149-151.

Fairely, K. F. (1970). Cephalexin in recurrent urinary tract infection.Postgrad. med. J., 46, Suppl. (Oct.), 24-27.

Fass, R. J., Perkins, R. L., and Saslow, S. (1970). Positive directCoombs tests associated with cephaloridine therapy. J. Amer.med. Ass., 213, 121-123.

Foord, R. D. (1969). Cephaloridine and the kidney. In Progress inAntimicrobial and Anticancer Chemotherapy. Proceedings of the6th International Congress ofChemotherapy, Tokyo 1969, vol. 1,pp. 597-604.

Susannah Eykyn428

Use and control of cephalosporins

Gralnick, H. R., Wright, L. D., Jr., and McGinniss, M. H. (1967).Coombs positive reactions associated with sodium cephalothintherapy. J. Amer. med. Ass., 199, 725-726.

Kabins, S. A., Kelner, B., Walton, E., and Goldstein, E. (1970).Cephalexin therapy as related to renal function. Amer. J. med.Sci., 259, 133-142.

Kind, A. C., Kestle, D. G., Standiford, H. C., Freeman, P., andKirby, W. M. M. (1969a). Development of Staphylococci cross-resistant to cephalexin and methicillin. Antimicrob. AgentsChemother., 1968, 8, 405-409.

Kind, A. C., Kestle, D. G., Standiford, H. C., and Kirby, W. M. M.(1968). Laboratory and clinical experience with cephalexin.Antimicrob. Agents Chemother., 1968, 8, 361-365.

Kunin, C. M., and Atuk, N. (1966). Excretion of cephaloridine andcephalothin in patients with renal impairment. New Engl. J.Med., 274, 4, 654-656.

Lawson, D. H., Macadam, R. F., Singh, H., Gavras, H., and Linton,A. L. (1970). The nephrotoxity of cephaloridine. Postgrad.med. J., 46, Suppl. (Oct.), 36-39.

Leigh, D. A., Faiers, M. C., and Brumfitt, W. (1970). Laboratory andclinical studies with cephalexin. Postgrad. med. J., 46, Suppl.(Oct.), 69-74.

429

Molthan, L., Reidenberg, M. M., and Eichman, M. F. (1967). Positivedirect Coombs tests due to cephalothin. New Engi. J. Med., 277,3, 123-125.

Perkins, R. L., Apicella, M. A., Lee, I.-S., Cuppage, F. E., andSaslaw, S. (1968). Cephaloridine and cephalothin: comparativestudies of potential nephrotoxicity. J. Lab. clin. Med., 71,75-84.

Pickering, M. J., Spooner, G. R., de Quesada, A., and Cade, J. R.(1970). Declining renal function associated with administrationof cephalothin. Sth. med. J., 63, 426-428.

Rahal, J. J., Jr., Meyers, B. R., and Weinstein, L. (1968). Treatment ofbacterial endocarditis with cephalothin. New Engl. J. Med.,279, 4, 1305-1308.

Turcotte, J. G., Herrmann, T. J., Haig, O., O'Dell, C., Jr., Cerny, J.,and Greene, J. A.,Jr. (1970). Oliguria following transplantationof kidneys donated by blood relatives. Surg., Gynec. andObstet., 130, 29-36.

Weinstein, L., and Kaplan, K. (1970). The cephalosporins: micro-biological, chemical and pharmacological properties and use inchemotherapy of infection. Review. Ann. intern. Med., 72,729-739.