vancomycin ototoxicity and nephrotoxicity

Contents

Summary

Adverse Drug Experience Review

Medical Toxicology 3: 376-386 (1988)0112-5966/88/0009-0376/$05.50/0© ADIS Press LimitedAll rights reserved.

Vancomycin Ototoxicity and NephrotoxicityA Review

George R. Bailie and Deborah NealUniversity of Manchester Department of Pharmacy, Hope Hospital, Salford, andDepartment of Pharmacy, Birch Hill Hospital, Rochdale, England

Summary 3761. Pharmacokinetics 3772. Ototoxicity 3773. Nephrotoxicity 3814. Conclusion 384

Vancomycin has been in clinical use as a potent antistaphylococcal antibiotic for over30 years. Most reports ofototoxicity and nephrotoxicity have been associated with early,relatively impure, formulations of vancomycin. This paper reviews the literature concerning vancomycin otoxocity and nephrotoxicity and the evidence for their correlation withthe therapeutic serum concentration range.

There have been 28 reports ofvancomycin-associated ototoxicity published in the medical literature since 1958. It remains unclear whether any diminution in hearing is permanent or reversible. Few patients in the literature had follow-up audiometry and thehearing impairment tends to be at higher frequencies. Several authors reported peak serumvancomycin concentrations, but the exact time these were drawn with respect to the lastdose is mostly unclear. In other reports, the 'peak ' concentrations noted 3 to 6 hours afterthe last dose are probably indicative of much higher concentrations because of vancomycin's rapid phase ofdistribution.

More than half the 57 cases of reported nephrotoxicity due to vancomycin occurredwithin the first 6 years of the drug's use. Many of these patients also had pre-existingrenal dysfunction or were concomitantly receiving other nephrotoxic agents. It is unclearwhether the coadministration ofaminoglycosides produces a synergistic toxicity. The exactincidence of nephrotoxicity is uncertain, but is probably less with the current, relativelypure, product. The correlation of nephrotoxicity with certain serum vancomycin concentrations remains to be clarified.

Other aspects also require clarification, such as when to draw samples to determinepeak serum concentrations and whether or not routine measurements are necessary at all.

In the absence ofbetter guidelines, efforts should be made to tailor individual patient'sregimens to produce peak and trough serum vancomycin concentrations to within thewidely accepted ranges of30 to 40 and 5 to 10 mg/L, respectively. In addition, the concomitant use ofother potentially nephrotoxic and ototoxic agents should be avoided.

Vancomycin Ototoxicity and Nephrotoxicity

Vancomycin is a bactericidal glycopeptide antibiotic which was introduced for clinical use in 1956.Initially vancomycin was known as 'MississippiMud' because early preparations of the drug contained substantial quantities of impurities. At thattime, vancomycin was extracted using picric acidprecipitations and was only about 70% pure (dataon file, Lilly Research Laboratories). Later, in theI960s, the manufacturers used an ion-exchangeresin resulting in the formation of crystalline copper-complex vancomycin which increased purityto about 75%. In 1985, a phosphate manufacturingmethod was used which further increased purity toabout 80% and finally, this year, a high performance liquid chromatographic method has been implemented which can produce vancomycin with apurity of 92 to 95%; according to the manufacturers this increased purity should be associated withfewer adverse effects. Vancomycin was readily accepted in the late 1950s, but reports of ototoxicityand nephrotoxicity and the introduction of antistaphylococcal penicillins and cephalosporins curtailed its widespread use. However, emergingstrainsof methicillin-resistant Staphylococcus aureus ensured that vancomycin be re-evaluated, and its useis now undergqing a renaissance.

In over 30 years of use, vancomycin has beenbranded as a toxic agent. Although the toxicity hasbeen well publicised, the actual number of patientsexperiencing ototoxicity and nephrotoxicity is quitesmall. The objective of this paper therefore is toreview the literature describing the reports of ototoxicity and nephrotoxicity and to relate these toserum vancomycin concentrations where possible.

The reader is referred elsewhere for detailedgeneral reviews of the pharmacology, clinical useand pharmacokinetics of vancomycin (Cheung &

DiPiro 1986; Matzke et al 1986). Other toxicitiessuch as cutaneous reactions, thrombophlebitis, redneck syndrome (Pau & Khakoo 1983), fits (Bailieet al. 1985), cardiac arrest (Mayhew & Deutsch1985), hypotension (Southorn et al. 1986)and neutropenia (Koo et al 1986) will not be discussed inthis review.

377

1. Pharmacokinetics

Mono- (Matzke et al. 1987), bi- (Rotschafer etal. 1982) and and triexponential (Moellering et al.1981a)concentration versus time profiles have beendescribed for vancomycin in patients with normalrenal function , and mono- (Matzke et al. 1984) orbiexponential (Cunha & Ristuccia 1983) profiles inpatients with renal impairment. Thus, there is apronounced distribution or a phase, during whichthere is also some elimination. Following completion of distribution [which has a half-life of about30 minutes (Rotschafer et al. 1982)], there is anelimination or {3 phase, which has a half-life inadults of up to 13 hours (Rotschafer et al. 1982).

The a phase does not appear to be significantlyaltered in patients with renal impairment. However, the elimination half-life increases in proportion to the decline in creatinine clearance (Moellering et al. 1981b).

2. Ototoxicity

Ototoxicity is widely considered to be one ofvancomycin's major systemic side effects. The exact mechanism for the production of ototoxicity byvancomycin is unclear, but various possible methods for any ototoxic agent have been reviewed(0'Arcy & Griffon 1979). The ototoxicity associated with vancomycin is characterised by damageto the auditory nerve, initially affecting high frequency sensory hairs in the cochlea, then middleand low frequency hairs, and eventually leading toa complete hearing loss. Thus, high tone deafnessoccurs before total deafness in all frequencies. Haircell degeneration is irreversible and the deafnessproduced is therefore permanent. Loss of acuity tohigh frequency sound and tinnitus are frequentclinical antecedents to deafness due to vancomycin(Fekety 1982), and it is recommended that tinnitusbe regarded as a sign to discontinue treatment(Reynolds 1982).

Although ototoxicity due to vancomycin hasbeen well publicised, the actual number of patientswho have experienced this adverse effect is quitesmall. Reviews of the literature from 1956 to 1986


revealed that vancomycin has been reported asbeing implicated as causing ototoxicity in 28patients. It is quite possible that there are unreported cases.

It is difficult to determine the exact potentialototoxicity of vancomycin from early reports. Manyof these patients were receiving concomitantaminoglycosides (Dangerfield et al. 1960; Dutton& Elmes 1959; Hook & Johnson 1978; Kirby et al.1960) [frequently streptomycin or neomycin, whichare themselves potentially highly ototoxic], had preexisting renal disease (Dangerfield et al. 1960; Dutton & Elmes 1959; Geraci et al. 1958; Waisbren etal. 1959-1960; Woodley & Hall 1961) and/or lifethreatening staphylococcal infections with accompanying failure of a major organ system, such ascongestive cardiac failure (Geraci et al. 1958) [tableI]. Furthermore, assessment of ototoxicity in thesereports has tended to be subjective, since mostpatients did not undergo audiometric testing during the course of therapy. Thus, hearing deficien-

378

cies were not reported unless they were substantialand at lower frequencies within the hearing range(2,000 to 3,OOOHz). Also, early formulations ofvancomycin contained substantial quantities ofimpurities which may themselves have had an ototoxic potential.

Hook and Johnson (1978) reported a series of15 patients receiving vancomycin. 10 patients hadserial audiograms undertaken. Three of thesepatients had a high frequency hearing loss (above4,OOOHz), although there was no detectable impairment in social hearing. The peak serum vancomycin concentrations were recorded (table I), although the timing of these with respect to doseswas not documented. The total vancomycin doseof these patients was 1 to 2 g/day for 1 to 6 weeks.

It was soon realised that patients with severerenal impairment were highly susceptible to the development of ototoxicity, especially if doses ofvancomycin were administered which were normalfor patients without renal dysfunction. The highest

Table J. Case reports from the medical literature describing vancomycin ototoxicity

Reference No. of patients Serum Time after Contributing factorsconcentration infusion(mg/L) (h)

Hook & Johnson (1978) 3/15 25 ? t Serum creatinine, 3 pts25 ? Streptomycin, 2pts

50 ?Dutton & Elmes (1959) 5/9 ? ? t Blood urea nitrogen, 4 pts

Streptomycin, 1 ptDangerfield et al. (1960) 3 ? ? Kanamycin, 1 pt

Kirby et al. (1960) 1 ? ? 1 month neomycin with some

ototoxicity

Geraci et al. (1958) 2 80-100 3-6 t Blood urea nitrogen, shock 1 pt

Woodley & Hall (1961) 2 ? ? t Blood urea nitrogen

Waisbren et al. (1959-1960) 1 ? ? t Blood urea nitrogen

Ahmad et at, (1982) 1 P 37 ? Haemodialysis

Traber & Levine (1981) 1 22-50 1

Mellor et al. (1984) 3 P < 25 ?Mellor et al. (1985) 4/34 P 25, T 19 20 min Amlkacin 1 pt

Gentamicin, steroids , cytotoxics

Louria et al. (1961) ? ? Novobiocin, chloramphenicol,

systemic lupus erythematosus,previous hearing loss

Morris & Bilinsky (1971) ? ? Previous hearing loss

Abbreviations: P = peak; T = trough ; t = elevated; pt = patient.


incidence of ototoxicity was reported in 1959, when5 of 9 patients receiving a course of vancomycindeveloped ototoxicity (Dutton & Elmes 1959). All5 patients had severe impairment of renal function . Four of these 5 had blood urea nitrogen (BUN)concentrations greater than 300 mg/dl (50 mrnol/L) at the start of their course of treatment, but received standard doses (I to 2 gJday) ofvancomycinfor extended periods. One other patient in this series had a BUN of380 mg/dl (63.3 mmol/L) fallingto 280 mg/dl (46.7 mmol/L), but the effect on herhearing was not known. One patient had recentlyreceived streptomycin. The authors acknowledgedthat they were not aware of the significant potential for ototoxicity and the need for dose modification in patients with renal failure, and thereforeroutine serum vancomycin assays were not undertaken. One can only speculate at the magnitude ofthe trough serum vancomycin concentrations,which could have been markedly elevated.

Relating ototoxicity to vancomycin serum concentrations is difficult. Most reports have not determined serum concentrations of vancomycin.When concentrations were stated, there was oftenno documentation of the time samples were drawnwith respect to the last dose. Despite this , it is generally considered that ototoxicity is associated withserum vancomycin concentrations in the range 80to 100 mg/L (Kirby et al. 1960). This assumption,however, is based on data of only 2 patients, reported in 1958, and only I of these patients wasfully described in the paper (Geraci et al. 1958).This patient had a previous history of renal disease, a BUN of 76 mg/dl (12.7 mmol/L) on admission to hospital, and an unrecordable diastolicblood pressure. He was treated with vancomycin500mg 6-hourly, and slight tinnitus and deafnesswere noted on day 12. Renal function improvedslightly at first but by the end of treatment BUNwas elevated to concentrations of 80 to 100 mg/dl(13.3 to 16.7 mmol/L), Serum vancomycin concentrations were noted to be 95 mg/L at 3 hoursafter the dose and 80 mg/L at 6 hours. Because ofthe very marked a phase of distribution followingadministration of vancomycin, it is only possibleto speculate concerning the true peak concentra-

379

tions in this patient. Clearly it would have beensubstantially higher than 100 mg/L,

Recently, further attempts have been made todelineate the relationship between serum concentration and ototoxicity, including reports of patientswho have exhibited ototoxicity with normal andabnormal renal function (Ahmad et al. 1982; Hook& Johnson 1978; Mellor et al. 1984; Sorrell et al.1982; Traber & Levine 1981). Nine patients in totalhave been reported, of whom 7 had serum vancomycin concentrations monitored; peak concentrations ranged from 25 to 50 mg/L, and troughsobtained just before the next scheduled dose rangedfrom 13 to 32 mg/L, However, again, most authorsdid not specify when 'peak' concentrations wereobtained in relation to the dose. One patient withnormal renal function experienced ototoxicity withI-hour postinfusion concentrations of between 21.6and 49.2 mgJL (Traber & Levine 1981). However,a serum concentration determined at I hour afterinfusion probably does not adequately reflect thetrue peak concentration because of the marked adistribution phase. In these 9 subjects there was noclinical deficit in auditory acuity, and in 5 subjectstinnitus was reversible and high frequency hearingloss (above 4,OOOHz) either stabilised or improvedwith a reduction in vancomycin dose.

These data may suggest a causal relationship between vancomycin administration, vancomycinserum concentration and auditory toxicity. However, they do not suggest that mild ototoxicity maybe reversed or prevented by close monitoring ofvancomycin serum concentrations. It is not clearwhether excess peak or trough concentrations areprimarily related to the development of ototoxicity. Indeed, there is still considerable controversyas to when samples should be drawn to estimatepeak concentrations. Recommendations have beenmade to draw samples at 15 minutes (Lake & Peterson 1985; Rotschafer et al. 1982), 60 minutes(Healy et al. 1987) or more than 3 hours (Matzkeet al. 1984) after a l-hour infusion. There appearsto be no overwhelming evidence in favour of anyone of these recommendations.

In a recent prospective study (Mellor et al, 1985),2 of 34 patients complained of tinnitus and diz-


ziness. One of these was a 20-year-old male whowas receiving no other drugs. This patient had peak(drawn 20 minutes after the infusion) and troughserum vancomycin concentrations of25 and 19 mg/L, respectively. The second patient also had normal renal function, but was receiving amikacinconcomitantly, so serum vancomycin concentrations were not obtained. The symptoms resolvedfor both these patients within 24 hours of the vancomycin being discontinued. Complete audiometric tests were then completed for some of the 34patients. Acute hearing loss was noted in 1 otherpatient who had a 30dB loss in one ear at 6,000Hz.This patient had normal renal function. Follow-upwas unavailable. One further patient had a delayedonset hearing, loss. Threshold increased by 20dB at250, 500 and 4,000Hz in the left ear and by 20dBat 500Hz in the right ear. Serum vancomycin concentrations were unavailable for these patients. Itis interesting to note that youth and 'acceptable'serum concentrations did not provide protectionagainst ototoxicity in this series, and that the authors concluded by stating they noticed no evidenceof synergistic toxicity between vancomycin and theaminoglycosides.

One report suggested ototoxicity in a 23-yearold White male with disseminated lupus erythematosus with Staphylococcus aureus endocarditis(Louria et al. 1961). The hearing loss began whilethe patient was receiving chloramphenicol and novobiocin several months after completing a courseof vancomycin, and was progressive over a periodof 4 months. Vancomycin, 3g daily for 6 weeks,was subsequently administered, again with a considerable worsening of his hearing impairment. Itis unclear whether this condition resolved.

A 4-year-old boy undergoing continuous ambulatory peritoneal dialysis accidentally received 6loading doses of 185mg of vancomycin, resultingin a plasma concentration of 121 mg/L (Hekster etal. 1986). He was treated with 2-hourly continuousambulatory peritoneal dialysis exchanges and, despite an elimination half-life of 21 hours, showedno evidence of ototoxicity on audiometric testing.

Another boy on haemodialysis received 2 dosesof vancomycin in error. This resulted in a 'peak'

380

serum concentration of 92.5 mg/L, but there wasno ototoxicity documented by brain stem-evokedresponses and behavioural audiometry (Schaad etal. 1981). It was unclear when the peak serum sample was taken.

In another report, no detectable ototoxicity onpure tone audiometry testing was found among 12patients on long term haemodialysis who receivedvancomycin for Staphylococcus aureus bacteraemia (Masur et al. 1983). Serum concentrationswere obtained 45 to 90 minutes after a 30-minuteinfusion and ranged from 5.5 to 40 mg/L, Therewere also no clinically apparent cases reportedamong 100 treatment courses in 98 patients in aretrospective study by Farber and Moellering(1983).

There is 1 additional report of possible ototoxicity which subsequently resolved completely(Morris & Bilinsky 1981), but no serum concentrations were recorded. This patient developed abilateral retinal artery occlusion and perceptivedeafness following a bilateral nephrectomy foraccelerated hypertension while receiving vancomycin. Other patients in this series had serialaudiometric studies which remained stable on vancomycin treatment.

Even almost a decade ago, the ototoxic potential of vancomycin was a matter of controversy.Friedlander (1979) suggested that vancomycin cameeleventh out of 15 in decreasing order of toxicitycompared with other known ototoxic agents. Hesuggested that its major contribution was cochleotoxicity as opposed to vestibular toxicity. Of noteis the fact that ototoxicity due to orally administered vancomycin appears not to have been reported (Kavanagh & McCabe 1983). In summary,therefore, vancomycin ototoxicity, manifested byirreversible deafness, is not as common as is perpetuated in the general literature. The true incidence is unknown, but is quite small. It is thereforethe opinion of the authors that ototoxicity does notconstitute a common side effect of vancomycin.

3. Nephrotoxicity

57 cases of vancomycin-associated nephrotoxicity have been reported in the literature, over 50%

of which appeared during the first 6 years of the


drug's use (Crumley, personal communication;Dangerfield et al. 1960; Dutton & Elmes 1959; Eisenberg et al. 1981; Farber & Moellering 1983;Hook & Johnson 1978; Kirby & Divelbiss 19561957; Kirby et al. 1960; Spears & Koch 1959-1960;Waisbren et al. 1959-1960; Woodley & Hall 1961)[table II]. Earlier, less purified, preparations mayhave had more nephrotoxic potential than today'spreparation (data on file, Lilly Research Laboratories). The early assessments were based on observations made after the completion of therapyrather than as a part of the prospective evaluationof the drug. Thus, the precise incidence of nephrotoxicity during the 1950s and 1960s is difficultto document.

Some early reports were relatively nonspecific.

381

Dutton and Elmes (1959) described the effects onrenal function in a series of 9 patients receivingvancomycin. Four patients with pre-existing renaldysfunction received total doses of vancomycin of6 to 13g over short periods (2 to 5 days), resultingin high BUN concentrations in all, and each of thesepatients became clinically deaf. All patients subsequently died, although not necessarily as a directresult of high serum vancomycin concentrations.Serum vancomycin concentrations were not reported for these patients. However, the methods ofvancomycin administration were clearly defined bythese authors. One method of administration involved the direct injection of vancomycin in 20mlof saline over a 5-minute period. It is quite likelythis procedure could result in high peak concen-

Table II. Case reports from the medical literature describing Vancomycin nephrotoxicity

Reference No. of patients Serum Time after Contributing factors

concentration infusion

(mg/L) (h)

Dutton & Elmes (1959) 4/9 ? ? Pre-existing renal disease

Dangerfield et al. (1960) 11/85 ? ? t Serum creatinine

Kirby et al. (1960) 0/33 ? ? t Blood urea nitrogen, urinary

sediment

Woodley & Hall (1961) 1/25 ? ? t Blood urea nitrogen

Waisbren et al. (1959-1960) 3/54 ? ? t Blood urea nitrogen

Mellor et al. (1985) 3/34 acute ? 20 min Gentamicin, frusemide, 1 pt:

t Serum creatinine

3/34 late ? ?Farber & Moellering (1983) 12/348 See text t Serum creatinine

3/60b

Crumley (personal 2 ? ?communication)

Eisenberg et al. (1981) 1 P< 30 ? Interstitial nephritis

Kirby & Divelbiss (1956-1957) 2 ? ? 2g doses

Spears & Koch (1959-1960) 1/53 ? ?Cafferkey et al. (1982) 1/23 ? ? t Serum creatinine, 2 previous

courses of gentamicin

Odio et al. (1984) 4 ? ? Concurrent aminoglycosides, 3

patients had pre-existing renal

disease

Dean et al. (1985) 2/19b ? ?

2/98 ? ?Craven et al. (1983) 2 ? ? t Serum creatinine

a Vancomycin and aminoglycoside concurrently.

b Vancomycin alone.

Abbreviations: see table I.


trations. Unfortunately, the paper does not state bywhich method the nephrotoxic patients receivedvancomycin.

Dangerfield et al. (1960) noted nephrotoxicityin 11 of 85 patients in their series. They definednephrotoxicity as elevations of 0.5 mg/dl (45.5Jl,mol/L) in serum creatinine which were otherwiseunexplained. Eight of these patients had no evidence of pre-existing renal dysfunction. In these 8patients, follow-up serum creatinines were obtained which demonstrated a return to baselinerenal function within 3 to 4 weeks. Three of the 11patients had pre-existing azotaemia. One of thesehad a return to baseline renal function, and theremaining 2 died of renal failure. No serum vancomycin concentrations were reported.

Kirby and Divelbiss (1956-1957) noted granularcasts in the urine of 2 healthy volunteers, 4 to 8hours after the intravenous administration of 2gdoses of vancomycin over 10 to 20 minutes. Oneof the 2 ran a low grade fever for 24 hours and hisBUN rose from 14 mg/dl (2.3 mmol/L) predose to20 mg/dl (3.3 mmol/L) a day later. Some patientstreated with vancomycin also had suggestive signsof 'renal irritation'. Two who died after severe toxicfebrile reactions showed brownish deposits in therenal tubules which it was suspected may have beenvancomycin, although the deposits were not investigated. However, there was no evidence of inflammation or tubular damage. There is no documentation of doses or duration of vancomycintreatment in these patients. The authors did statethat 1 other patient with pre-existing renal dysfunction received 2g vancomycin daily for 8 dayswith no further change in BUN or of findings ofurine sediment.

Others noted albuminuria and microscopic haematuria in a 1-year-old female patient (Spears &Koch 1959-1960). This patient was receiving 40 mg/kg/day of vancomycin intravenously for the treatment of a staphylococcal abscess and cellulitis. Theexact method of administration is unclear, sincethe authors attempted several methods of intravenous and intramuscular administration. Serumvancomycin concentrations were not reported forthis patient. The condition cleared spontaneously

382

on discontinuation of the antibiotic. No concomitant elevation of BUN was noted.

Many reported patients had similar compounding factors to those described in the ototoxicity literature: concomitant aminoglycosides, pre-existingrenal disease, or major organ system failure. Theseproblems could have caused impairment of renalfunction unrelated to vancomycin therapy and leadto an erroneous diagnosis of toxicity. Despite theseproblems, concern about nephrotoxicity was one ofthe main factors which limited the clinical use ofvancomycin.

The relationship of nephrotoxicity to vancomycin serum concentrations is difficult to determine. Most investigators did not measure serumvancomycin concentrations or, if they did, they didnot specify the timing of blood sampling in relationship to drug administration. Farber and Moellering (1983) studied a group of 98 patients receiving 100 vancomycin courses, of which 94 coursescould be evaluated for the development of nephrotoxicity. 34 patients received concomitantaminoglycosides and 60 received vancomycinalone. Nephrotoxicity was defined as a rise of atleast 0.5 mg/dl (45.5 Jl,mol/L) in serum creatinineabove a pretreatment baseline. 12 of the 34 patientsreceiving concomitant aminoglycosides (35%) developed nephrotoxicity and 3 of 60 patients (5%)receiving vancomycin alone developed this toxicity. The mean age and duration of vancomycintreatment was greater in patients developing nephrotoxicity. Serum vancomycin concentrations weredocumented for many of the 15 nephrotoxicpatients, but the timing ofwhen peak samples weredrawn is not recorded. The 3 nephrotoxic patientswho received vancomycin without aminoglycosides had elevated trough concentrations of 39, 48and 65 mg/L, Of all nephrotoxic patients, all except 1 had 'peak' vancomycin concentrations in therange 19 to 49 mg/L, The last patient (with a troughof 65 mg/L) had a 'peak' concentration of 76 mg/L. Although derived from small numbers ofpatients, this appears to be the only basis for a correlation between vancomycin serum concentrationand nephrotoxicity.

More recently, systematic assessments were car-


ried out to ascertain the nephrotoxic potential ofvancomycin in rats (Aronoff et al. 1981; Wold &Turnipseed 1981). The importance of these studiesis unclear since although the same animal modelhas been used to determine the nephrotoxic potential of the various aminoglycoside antibiotics, itmay not be entirely appropriate to extrapolate animal data to the clinical situation. Arnoffet al. (1981)evaluated renal function and histological changesin rats given repeated doses of vancomycin of upto 400 rug/kg/day. Creatinine clearance, urinaryprotein excretion and urine osmolality did not alter significantly between control and treated groups.Although mild histological changes were seen inrats given 400 rug/kg/day, the authors concludedthat vancomycin did not alter renal function.

Wold and Turnipseed (1981) evaluated thenephrotoxic potential of vancomycin in rats afterdoses of 75, 150 and 350 rug/kg/day given aloneand in combination with tobramycin 60 mg/kg, Nochange in renal function was observed after administration of vancomycin alone in doses of 75 or 150mg/kg, or tobramycin alone. However, significantdecline in renal function and increases in kidneyweight were seen in animals given vancomycin 350rug/kg alone or 150 mg/kg in combination withtobramycin. It was concluded that concomitantadministration of vancomycin and tobramycin torats resulted in significantly increased nephrotoxicity compared to that caused by either agent alone.The results were confirmed in humans by Farberand Moellering's (1983) retrospective study whichstated that although this high incidence could haveresulted from bias in selection of the patients or toaminoglycoside therapy alone, the possibility ofadditive toxicity between vancomycin and theaminoglycosides must be considered.

Serum creatinine concentrations rose sharplywith vancomycin therapy in 1 of a group of 23patients reported by Cafferkey et al. (1982). Thispatient's pretherapy creatinine concentrations hadbeen 'high normal'. This patient had also receivedgentamicin for two 3-week periods in the previous3 months. It was considered that renal impairmentwas present prior to vancomycin therapy, and ittherefore remains difficult to assess whether van-

383

comycin had aggravated the renal damage. Theauthors stated that vancomycin concentrations indicated reduced clearance of the drug, but the actual concentrations were not reported. They concluded that drug toxicity was not a serious problem.

Nephrotoxicity, defined as a 2-fold or greaterincrease in serum creatinine concentration, occurred in 4 children given vancomycin and anaminoglycoside (Odio et al. 1984). Vancomycin wasgiven intravenously in doses of 40 rug/kg/day in 4divided doses to each child, but serum vancomycinconcentrations were never determined. Thesepatients received combined treatment for 3 to 31days.

Vancomycin-induced nephrotoxicity was seenin 2 of 19 (11oro) children receiving vancomycinalone, and in 2 of 9 (22%) who received at least 1week of concurrent vancomycin and aminoglycoside therapy (Dean et al. 1985). Serum creatinineconcentrations returned to baseline when vancomycin serum concentrations were adjusted to obtain peak and trough concentrations of 20 to 40mg/L and 5 to 10 mg/L, respectively. It was notstated when peak samples were obtained. It wassuggested in this report that pharmacokineticmonitoring of doses in children should avert, minimise or reverse nephrotoxicity during therapy. Thiswas also suggested in a study which confirmed thatpaediatric burn patients also had considerable variations in their vancomycin dosage requirements(Bailie et al. 1984). The authors of this study foundthat doses of up to 60 rug/kg/day of vancomycinwere required to maintain these patients' peak andtrough serum vancomycin concentrations in theranges of 30 to 40 mg/L and 5 to 10 mg/L, respectively. They aimed for peak concentrations,drawn 15 minutes after a l-hour infusion, of 25mg/L, to ensure that the true peak was 30 to 40mg/L,

In a recent prospective study of vancomycintoxicity in 34 patients, the rate of acute nephrotoxicity was about 7°ro (3 out of 39 courses) and therate of delayed toxicity after the termination oftherapy was 9% (Mellor et al. 1985). These authorsdefined nephrotoxicity as a rise in serum creatinineof 0.5 mg/dl (45.5 ~mol/L) if the initial creatinine


was 3 mg/dl (272.7 ~mol/L) or less, or a rise ofgreater than 1.0 mg/dl (90.9 ~mol/L) if the initialcreatinine was greater than 3 mg/dl (272.7 ~mol/

L). Serum creatinine concentrations were measured for up to I week after vancomycin treatmentfor acute toxicity, and up to 4 weeks for delayedtoxicity. The authors stated the peak vancomycinconcentrations were obtained 20 minutes after thedose, which again demonstrates the controversysurrounding the timing of drawing samples formeasurement of peak concentrations. No individual serum vancomycin concentrations were reported. These authors saw no evidence of synergistic toxicity between vancomycin andaminoglycosides, although many of their patientshad serious underlying diseases which may haveaffected serum creatinine concentrations. All 3patients with acute nephrotoxicity died, althoughthis was not directly attributed to high serum vancomycin concentrations. They had been receivingvancomycin doses of 24 to 30g for 12 to 15 days.Serum vancomycin concentrations were unavailable for these 3 patients because of multiple antibiotic administration.

There has been I case report of interstitial nephritis in a patient receiving vancomycin for rightsided endocarditis and pneumonia, secondary tointravenous drug abuse. It was characterised byazotaemia, fever and eosinophiluria and resolvedwhen the drug was withdrawn (Eisenberg et al.1981). 'Peak' serum concentrations were maintained below 30 mgfL during therapy.

Craven et al. (1983)stated that in 2 patients withcompromised renal function prior to vancomycintherapy, subsequent elevations of serum creatinineconcentrations may have been caused by vancomycin. No further details were provided.

In summary, therefore, it would appear thatthere is a certain degree of nephrotoxicity inducedby vancomycin. Evidence would suggest that therisk increaseswith higher serum concentrations, butthere are no firm data for defining a therapeuticrange for vancomycin. It is not possible to statethe incidence of nephrotoxicity.

384

4. Conclusion

There does not appear to be any evidence correlating the rate of eradication of infection to specific vancomycin serum concentrations. It wouldappear that the correlation of vancomycin-inducedototoxicity and nephrotoxicity with certain serumconcentrations also remains to be clarified. Indeed,the controversy continues, with advocates both forand against the routine measurement of serumvancomycin concentrations as an indicator of toxicity as well as efficacy (Edwards & Pancorbo 1987;Rodvold et al. 1987;Sayers & Shimasaki 1988). Itmay be that the magnitude of the peak serum vancomycin concentration per se is relatively unimportant. Rather, the total area under the serumconcentration-time curve (AVC) might be a moreaccurate indicator of impending toxicity (Keller1984).

Until this correlation is clarified, the controversy will continue. It would appear that in theabsence of better guidelines, efforts should be madeto tailor individual patients' vancomycin regimensto provide vancomycin concentrations within thecurrently widely accepted, if somewhat arbitrary,ranges of 30 to 40 mg/L for peak and 5 to 10 mg/L for trough (Geraci 1977). The use of vancomycinin patients who are deemed to be at risk for thedevelopment of toxicity should be minimised, asshould the concomitant administration of otherpotentially ototoxic and nephrotoxic drugs.

References

Ahmad R, Raichura N, Kilbane V, Whitfield E. Vancomycin: areappraisal.Correspondence. British Medical Journal 284: 19531954, 1982

Aronoff GR, Sloan RS, DinwiddIe CB, et al. Effects of vancomycin on renal function in rats. Antimicrobial Agents andChemotherapy 19: 306-308, 1981

Bailie GR, Ackerman BH, Fischer J, et al. Increased vancom ycindosage requirements in young bum patients. Journal of BumCare and Rehabilitation 5: 376-378; 1984

Bailie GR, Yu R, Morton R, Waldek S. Vancomycin, red-necksyndrome and fits. Lancet 2: 279-280, 1985

Cafferkey MT, Hone R, Keane T. Severe staphylococcal infections treated with vancomycin . Journal of AntimicrobialChemotherapy 9: 69-74; 19li2

Cheung RPF, DiPiro JT. Vancomycin: an update. Pharmacotherapy 6: 153-169, 1986


Cra ven DE, Kollisch NR , Hsieh CR, Connolly Jr MG, McCabeWR. Vancomycin treatment of bacteraemia caused by oxacillin-res istant Staphylococcus aureus: comparisons with B-Iactam ant ibiot ic treatment of bacteraemia caused by oxacillinsensit ive Staphylococcusaureus. Journal ofInfectious Diseases147: 137-143, 1983

Cunha BA, Ristuccia AM. Clinical usefulness of vancomycin.Clin ical Pharmacy 2: 417-424, 1983

Dangerfield HC, Hewitt WL, Monzon OT, et al. Clinical useof vancomycin. Antimicrobial Agents Annual: 428-438,1960

D'Arcy PF, GritTon JP (Eds) , Iatrogenic diseases , 2nd ed. pp. 393400, Oxford University Press, Oxford , 1979

Dean RP, Wagner DJ, Tolpin MD. Vancomycin/arninoglycosidetoxicity . Journal of Paediatrics 106: 861-862, 1985

Dutton AAC, Elmes Pc. Vancomycin: report on treatment ofpatients with severe staphylococcal infections. British MedicalJournal I: 1144-1149, 1959

Edwards DJ, Pancorbo S. Routine monitoring of serum vancomycin concentrations: waiting for proof of its value . ClinicalPharmacy 6: 652-654, 1987

Eisenberg ES, Robbins N, Lenci M. Vancomycin and interstitialnephritis. Annals of Internal Medicine 95: 658, 1981

Farber BF, Moellering RC. Retrospective study of the toxicity ofpreparations of vancomycin from 1974-1981. AntimicrobialAgents and Chemotherapy 23: 138-141, 1983

Fekety R. Vancomycin. Medical Clinics of North America 66:175-181, 1982

Friedlander IR. Ototoxic drugs and the detection of ototoxicity.New England Journal of Medicine 301: 213-214, 1979

Geraci JE, Heilman FR, Nichols DR, Wellman WE. Antibiotictherapy of bacterial endocarditis (VII): vancomycin for acutemicrococcal endocarditis. Proceedings of the StatT Meetings ofthe Mayo Clin ic 33: 172-181, 1958

Geraci JE. Vancomycin. Mayo Clinic Proceedings 52: 631-634,1977

Healy DP, Polk RE, Garson ML, et al. Comparison of steadystate pharmacokinetics of two dosing regimens of vancomycinin normal volunteers. Ant imicrobial Agents and Chemotherapy 31: 393-397, 1987

Hekster YA, Vree TB, Weemaes CMR , Rotteveel JJ. Toxicologicand pharmacokinetic evaluation of a case of vancomycin intoxication during continuous ambulatory peritoneal dialysis.Pharmaceutisch Weekblad - Scientific Edition 8: 293-297,1986

Hook EW, Johnson ·WD. Vancomycin therapy of bacterialendocarditis. American Journal of Medicine 65: 411-415,1978

Kavanagh KT, MCCabe BF. Ototoxicity of oral neomycin andvancomycin. Laryngoscope 93: 649-653, 1983

Keller H. Cited in Dukes (Ed.) Meyler's Side etTectsofdrugs , 10thed. , p. 512, Elsevier, Amsterdam, 1984

Kirby WMM, Divelbiss CL. Vancomycin: clinical and laboratorystudies. Antibiotics Annual: 107-117,1956-1957

Kirby WMM , Perry DM, Bauer AW. Treatment ofstaphylococcalsept icaemia with vancomycin. New England Journal of Medicine 162: 49-55, 1960

Koo KB, Bachand RL, Chow AW. Vancomycin-induced neutropenia . Drug Intelligence and Clin ical Pharmacy 20: 780-782,1986

Lake KD, Peterson CD. A simplified dos ing method for init-

385

iating vancomycin therapy. Pharmacotherapy 5: 340-344,1985

Louria DB, Kaminski T, Buchman J. Vancomycin in severe staphylococcal infections. Archives of Internal Medicine 107: 125140,1961

Masur H, Francioli P, Ruddy M, Murray HW, Vancomycin serumlevels and toxicity in chronic haemodialysis patients with Staphylococcus aureusbacteremia. Clinical Nephrology 20: 85-88,1983

Matzke GR , McGory RW, Halstenson CE, Keane WF. Pharmacokinetics of vancomycin in patients with various degreesof renal function . Antimicrobal Agents and Chemotherapy 25:433-437, 1984

Matzke GR , Zhanel GG, Gua y DRP. Clinical pharmacokineticsof vancomycin. Clinical Pharmacokinetics II : 257-282, 1986

Mayhew JF, Deutsch S. Cardiac arrest following administrationof vancomycin. Canadian Anaesthetists Society Journal 32: 6566, 1985

Mellor JA, Kingdom J, CatTerkey M, Keane CT. Vancomycinototoxicity in patients with normal renal function. BritishJournal of Audiology 18: 179-180, 1984

Mellor JA, Kingdom J, CatTerkey M, Keane CT. Vancomycintoxicity : a prospective stud y. Journal ofAntimicrobial Chemotherapy 15: 773-780, 1985

Moellering RC, Krogstad DJ, Greenblatt DJ. Pharmacokineticsof vancomycin in normal subjects and in patients with reducedrenal function . Reviews of Infectious Disease 3 (Suppl .): 230235, 1981a

Moellering RC, Krogstad DJ, Greenblall DJ. Vancomycin therapy in patients with impaired renal funct ion: a nomogramfor dosage. Annals of Internal Medicine 94: 343-346,1981h

Morris AJ, Bilinsky RT. Prevention of staph ylococcal shunt infections by continuous vancomycin prophylaxis. AmericanJournal of the Medical Sciences 262: 87-92, 1971

Odio C, McCracken GH , Nelson JD . Nephrotoxicity associatedwith vancomycinjaminoglycoside therapy in four children.Journal of Paediatrics 105: 491-493, 1984

Pau AK, Khakoo R. "Red-neck syndrome" with slow infusion ofvancomycin. New England Journal of Medicine 313: 756-757,1983

Reynolds JEF (Ed). Martindale, the extra pharmacopoeia, 28thed., The Pharmaceutical Press, London, 1982

Rodvold KA, Zokufa H, Rotschafer Jc. Routine monitoring ofserum vancomycin concentrations: can waiting be justified?Clinical Pharmacy 6: 655-658, 1987

Rotschafer JC , Crossley K, Zaske DE, et al. Pharmacokinetics ofvancomycin; observations in 28 patients and dosage recommendations. Antimicrobial Agents and Chemotherapy 22: 391394, 1982

Sayers JFB, Shimasaki R. Rout ine monitoring of serum vancomycin concentrations: the answer lies in the middle. ClinicalPharmacy 7: 18, 1988

Schaad UB, Nelson JD , McCracken GH . Pharmacology and efficacy of vancomycin for staphylococcal infections in children.Review of Infectious Diseases 3: S282-S288, 198I

Sorrell TC, Packham DR, Shanker S, Folds M, Munro R.Vancomycin therapy for methicillin-resistant Staphylococcus aureus. Annals of Internal Medicine 97: 344-350,1982

Southorn PA, Plevak DJ, Wright AJ, Wilson WR. Adverse etTects

Van com ycin Ot otoxicity and Nephrotoxicit y

of vancomyc in admin istercd in the per ioperat ive per iod . MayoClinic Proceed ings 6 1: 72 1-724. 1986

Spears RL. Koch R. Th e usc of vancomycin in paed iatr ics. Ant ibiotics Annua l: 798-803. 1959-1960

Traber PG. Levine DP. Vanc omycin ototoxicity in a pati ent withnorm al renal funct ion . Annal s of Interna l Med icine 95: 458460. 1981

Waisbren BA. Kleinerman L. Skemp J. et al. Th e co mparativetoxicity and clinical effectiveness of vancomycin. ristocetin andkanam ycin . Antibi oti cs Annual: 497-515. 1959-1960

Wold JS. Turn ipseed SA. Toxicology of vancom ycin in labora-

386

to ry animals. Reviews of Infectious Diseases 3 (Suppl): 224229. 1981

Wood ley OW. Hall WH oThe treatment of severe staphylococcalinfectio ns with vancomyci n. Annals of Internal Medicine 55:235-249. 1961

Auth ors' address: Dr G.R. Bailie'. Associate Professor. Depart

ment of Pharmacy Practi ce. Alban y College of Pharmacy. 106

New Scotland Avenue . Alban y. NY 12208 (USA).

vancomycin ototoxicity and nephrotoxicity

Documents