Indian J Pediatr 1992; 59 : 213-219
Chloramphenicol Clearance in Typhoid Fever : Implications for Therapy Zulfiqar A Bhutta, Sarfaraz K Niazi*, and Amin Suria
Department of Pediatrics and Pharmacology, Faculty of Health Sciences, The Aga Khan University, Karachi, Pala~tan
We prospectively studied the pharmacokinetics of intravenous Chloramphenicol succinate (CS) in children (age 6 months-14 years) with culture proven typhoid fever (n = 30) and non typhoidal illnesses (n = 10). CS was administerecl in three different dosage regimens (50, 75 and 100 mg/kg/d-q 6 hourly). Uver function tests were monitored. Plasma trough and peak chloramphenicol concentrations were measured by HPLC analysis after 42 hrs. The 50 mg/kg/day dosage schedule was terminated midway through the study, as blood levels were consistently low and two patients with typhoid relapsed, children with typhoid had significantly lower clearance of CS in comparison with those with non-typhoidal illness (0.29 +_ 0.1 versus 0.5 + 0.37 1/ kg/hr, P 0.05). There was no significant difference between mean peak and trough concentrations of chloramphenicol on 100 mg/kg/day and 75 mg/ kg/day in children with typhoid. However, two children on 100 mg/kg/day dosage developed trough concentrations >20 mcg/ml. No correlation was found between CS clearance and serum bilirubin, SGPT (alanine transaminase) and alkaline phosphatase. Our data show altered clearance of CS in children with typhoid and suggests that 75 mg/kg/day may be a safer dose in children with hepatic dysfunction in typhoid.
Key Words : Chloramphenicol succinate (CS), Typhoid. Chloramphenicol is widely used in children in the treatment of meningitis caused by Ampicillin resistant strains of Hemophilus influenza.1 Despite the emergence of drug resistant strains and the availability of other Reprint requests : Dr. Zulfiqar A. Bhutta, Assistant Professor, Department of Pediatrics, Faculty of Health Sciences, The Aga Khan University,Karachi, Pakistan
antibiotics, chloramphenicol remains the treatment of choice in typhoid fever, for reasons of cost and ease of oral administration? Most of the organisms susceptible to chloramphenicol are inhibited by a serum concentration of the free drug between 5-20 mcg/ml? '4 Although a dosage range of 50100 mg/kg/day has been recommended in
213
214
T H E INDIAN J O U R N A L O F PEDIATRICS
older infants and children, 5,6 previous studies suggest that reliable levels may only be achieved with individual doses in excess of 22 mg/kg. 7,8The importance of appropriate dosing in children is highlights by observations that serum levels in excess of 25 pg/ml are associated with bone-marrow suppression9 and potentially toxic levels of chloramphenicol have been demonstrated in 41% of neonates and infants on therapy. 1~Although the "grey baby syndrome" in neonates has been classically associated with serum levels in excess of 50 mcg/ml, 11 serve and refractory acidosis has been recently described in children with even lower levels of chloramphenicol, tz~3 It is therefore necessary that adequate dosage criteria for therapy with chloramphenicol be established. The intravenous form, chloramphenicol succinate (CS) is hydrolysed in the liver to active chloramphenicol and a variable amount is lost in the urine) 4 The active chloramphenicol is mainly conjugated in the liver by glucuronidation prior to excretion through the bowel.15 It is therefore possible that alteration of liver function may significantly effect metabolism of chloramphenicol and indeed, potentially toxic levels have been demonstrated in some adults with liver impairment) 6,~7Typhoid fever has also been associated with significant hepatic dysfunction in a number of Studies in adults) 8,19 There is however, very little information on chloramphenicol metabolism and pharniacokinetics in typhoid fever in children, with or without concomitant hepatic dysfunction. We undertook a prospective study of chloramphenicol pharmacokinetics in children with typhoid fever, admitted to the Aga Khan University Hospital.
Vol. 59, No. 2
MATERIAL AND METHODS The protocol of the study was approved by the Human Subjects Protection Committee of AKUH. Patient selection. Over a 12 month period January-December 1988, all children admitted to the paediatric ward at A K U H with suspected typhoid fever (hyerpyrexia > hrs, abdominal tenderness and absence of localizing signs), were considered eligible for the study. Informed consent was obtained from the parents or guardians prior to inclusion in the study. The children were randomized after inclusion in the study to three dosage groups receiving 50, 75 and 100 mg/kg/day of IV CS respectively, in a blinded fashion. The drug was stopped at 72 hours if the results of blood and bone marrow cultures were negative. In culture positive cases, intravenous therapy was replaced by oral chloramphen|col palmitate when the patients became afebrile (temperature 37.5~ for 24 hours or more) and the treatment given for a total of 14 days. A precise daily record of clinical condition and six hourly vital signs including temperature, pulse and blood pressure was maintained. Investigations. After cultures, a baseline blood sample (3 cc) was obtained in all cases. The blood was analysed for a complete blood count and differential on a Coulter counter. Liver function tests and bicarbonate were estimated on an a~toanalyser (Beckman Inc, USA). These investigations were repeated with the 42 hours chloramphenicol levels. Chloramphenicol administration and assay. CS was administered in each dose group in six hourly doses by infusion over 30 minutes. A 3 ml sample of blood was ob-
215
BHU'ITA El" AL : CI ILORAMPHENICOL CLEARANCE WITH TYPHOID FEVER
tained just prior to the 8th dose i.e. at 42 hours for trough estimation. A 3 ml blood sample was drawn one hour after the end of infusion for estimation of peak chloramphenicol levels. Plasma was frozen at -80~ until analysis in fortnightly batches. Analysis was done, on 50 microlitres sample of plasma by high pressure liquid chromatography (HPLC) using the method of Neilsson Ehle et al, 19 with some modifications. (Novapak : microbondapak C 18 column, 25 cm * 2.1 mm id, Waters Associates, USA) was used. Mobile phase was acetic acid, water and methanol in ratios of 1 : 62 : 37. How rate of mobile phase was 1.0 ml per minute. The HPLC system (Waters Associates, USA) consisted of a pump M-(~30 a, a variable LC spectrophotometer 440 ~th absorbance set at 280 mm, WISP 710 B injector, Data Module 730 and a system controller 720. All solvents used were of ultra high purification HPLC grade. Free chloramphenicol for standards was obtained from Parke Davis (Pakistan) Ltd. Standard curve of chloramphenicol was constructed using 5,10,20,30 and 50 meg free chloramphenicol per ml. A four point standard curve, prepared in water, included c0ncentratios of 6.25 to 50 mcg/ml was made. Linear regression analysis of peak heights and chloramphenicol concentrati0ns gave a correlation coefficient of 0.99. Concentration of chloramphenicol in plasma was calculated from the standard curveand reported as mcg chloramphenlcol per ml. Since the analysis was done in fortnightly batches, no therapeutic alteration was made inany group on the basis of the levels estimated. Evaluation of chloramphenicol kinetics.
The chloramphenicol blood levels were fitted to an open single compartment model using an approximation that a steady state had been achieved before the eighth dose after 6-7 half lives had elapsed. The clearance was also calculated on the basis of the area under the curve (AUC) approximation. 2]'22 At steady state the total AUC is proportional to the clearance of the total dose given during the dosing interval, CL = Dose/AUC The AUC was calculated by A U C = ((Cpk + C48)/2)T Where T is the dosing interval, 6 hours. C48 wastaken as C42 or trough level (Figure 1). The basis for these calculations is expressed in Figure 1 and is similar to the approach used by Sack et al.21 Analysis of data was done using an integrated statistical and graphic package, Statgraphics (Statistical Graphics Corporation, USA). The t test was employed for comparison. RESULTS A total of 40 febrile children, admitted to the paediatric ward with suspected typhoid, were placed on CS. Of these 30 were subse-
m
mL~r~
mm J~
~ mB~
~m
~ K ~
'gl
II
4
i
It
i2
"
30
3e
~
48
' n n w In h m n .
Fig. 1.Calculation of pharmacokinetics parameters for steady state assumptions (see text for details)
216
TIlE INDIANJOURaNALOF PEDIATRICS
quently proven to have typhoid on cultures ~nd an alternative diganosis was made in he other ten. The various final diganosis made in the non-typhoid group were viral upper respiratory infection (4), bronchopneumonia (2), otitis media (2), varieella (1) and urinary tract infection (1). The characteristics of the patients in both groups are given in Table 1. The 50 mg/kg/day dosage of CS was stopped after the mid term analysis of the results of the study, as there were two relapses within three weeks of completion of a fourteen day course of therapy on this dosage, and because of concern about the observed low blood levels of chloramphenicol. Figure 2 shows the comparative peak and trough chloramphenicol concentrations for three dosage schedules in 30 children with typhoid. There was no significant difference in the peak (15.4 + 6.1 versus 19.9 _+ 12.2 mcg/ml, mean +_ SD) and trough (5.4 ~- 2.6 versus 8.8 • 7.7 mcg/ml) values on 75 and 100 mg/kg/day C,'S respectively. Three children on 100 mg/kg/day CS had peak chloramphenicol levels in excess of 309 mcg/ml. The mean trough and peak Table 1. Characteristics of the Patient (mean 95% CI) Typhoid N = 30
Non Typhoid n = 10
M.F. 21.9 7.3 Age (yr) 8.6 • 4.0 8.3 • 3.9 Wt (kg) 273 +_ 14.0 23.5 • 8.9 Hepatomegaly (%) 57% 10% Alanine transaminase 64.4 _+33.8 32.9 _+ 16.7 Ou/l) Alkaline phosphate 165.8 +- 53.5 172.1 -+ 66.5 Bicarbonate (mEq/1) 18.9 +_2.4 19.9 _+4.0
Vol. 59, No. 2
"'/'.'....")))7"?)))7')?Z'?)~'?/')I'Z'?)).'" l
A ~
B ~mOtk~d~
C ~00 nOmoSd~r
c~mplm~1~ Do~g, Fig. 2.Compa~son of peak chloramphenico! levels in three dosage groups in patients with
typhoid (Box-whisker plots). Data presented as median (horizontal bar) upper and lower quartiles (boxes) and range. Outliers represented as individal dots. levels of chloramphenicol on 50 mg/kg/day CS in children with typhoid were 4.3 _+ 2.5 and 8.5 • 1.1 mcg/ml respectively. These values were significantly lower than the peak (19.9 _+ i2.2 mcg/ml, p
Chloramphenicol Clearance in Typhoid Fever : Implications for Therapy Zulfiqar A Bhutta, Sarfaraz K Niazi*, and Amin Suria
Department of Pediatrics and Pharmacology, Faculty of Health Sciences, The Aga Khan University, Karachi, Pala~tan
We prospectively studied the pharmacokinetics of intravenous Chloramphenicol succinate (CS) in children (age 6 months-14 years) with culture proven typhoid fever (n = 30) and non typhoidal illnesses (n = 10). CS was administerecl in three different dosage regimens (50, 75 and 100 mg/kg/d-q 6 hourly). Uver function tests were monitored. Plasma trough and peak chloramphenicol concentrations were measured by HPLC analysis after 42 hrs. The 50 mg/kg/day dosage schedule was terminated midway through the study, as blood levels were consistently low and two patients with typhoid relapsed, children with typhoid had significantly lower clearance of CS in comparison with those with non-typhoidal illness (0.29 +_ 0.1 versus 0.5 + 0.37 1/ kg/hr, P 0.05). There was no significant difference between mean peak and trough concentrations of chloramphenicol on 100 mg/kg/day and 75 mg/ kg/day in children with typhoid. However, two children on 100 mg/kg/day dosage developed trough concentrations >20 mcg/ml. No correlation was found between CS clearance and serum bilirubin, SGPT (alanine transaminase) and alkaline phosphatase. Our data show altered clearance of CS in children with typhoid and suggests that 75 mg/kg/day may be a safer dose in children with hepatic dysfunction in typhoid.
Key Words : Chloramphenicol succinate (CS), Typhoid. Chloramphenicol is widely used in children in the treatment of meningitis caused by Ampicillin resistant strains of Hemophilus influenza.1 Despite the emergence of drug resistant strains and the availability of other Reprint requests : Dr. Zulfiqar A. Bhutta, Assistant Professor, Department of Pediatrics, Faculty of Health Sciences, The Aga Khan University,Karachi, Pakistan
antibiotics, chloramphenicol remains the treatment of choice in typhoid fever, for reasons of cost and ease of oral administration? Most of the organisms susceptible to chloramphenicol are inhibited by a serum concentration of the free drug between 5-20 mcg/ml? '4 Although a dosage range of 50100 mg/kg/day has been recommended in
213
214
T H E INDIAN J O U R N A L O F PEDIATRICS
older infants and children, 5,6 previous studies suggest that reliable levels may only be achieved with individual doses in excess of 22 mg/kg. 7,8The importance of appropriate dosing in children is highlights by observations that serum levels in excess of 25 pg/ml are associated with bone-marrow suppression9 and potentially toxic levels of chloramphenicol have been demonstrated in 41% of neonates and infants on therapy. 1~Although the "grey baby syndrome" in neonates has been classically associated with serum levels in excess of 50 mcg/ml, 11 serve and refractory acidosis has been recently described in children with even lower levels of chloramphenicol, tz~3 It is therefore necessary that adequate dosage criteria for therapy with chloramphenicol be established. The intravenous form, chloramphenicol succinate (CS) is hydrolysed in the liver to active chloramphenicol and a variable amount is lost in the urine) 4 The active chloramphenicol is mainly conjugated in the liver by glucuronidation prior to excretion through the bowel.15 It is therefore possible that alteration of liver function may significantly effect metabolism of chloramphenicol and indeed, potentially toxic levels have been demonstrated in some adults with liver impairment) 6,~7Typhoid fever has also been associated with significant hepatic dysfunction in a number of Studies in adults) 8,19 There is however, very little information on chloramphenicol metabolism and pharniacokinetics in typhoid fever in children, with or without concomitant hepatic dysfunction. We undertook a prospective study of chloramphenicol pharmacokinetics in children with typhoid fever, admitted to the Aga Khan University Hospital.
Vol. 59, No. 2
MATERIAL AND METHODS The protocol of the study was approved by the Human Subjects Protection Committee of AKUH. Patient selection. Over a 12 month period January-December 1988, all children admitted to the paediatric ward at A K U H with suspected typhoid fever (hyerpyrexia > hrs, abdominal tenderness and absence of localizing signs), were considered eligible for the study. Informed consent was obtained from the parents or guardians prior to inclusion in the study. The children were randomized after inclusion in the study to three dosage groups receiving 50, 75 and 100 mg/kg/day of IV CS respectively, in a blinded fashion. The drug was stopped at 72 hours if the results of blood and bone marrow cultures were negative. In culture positive cases, intravenous therapy was replaced by oral chloramphen|col palmitate when the patients became afebrile (temperature 37.5~ for 24 hours or more) and the treatment given for a total of 14 days. A precise daily record of clinical condition and six hourly vital signs including temperature, pulse and blood pressure was maintained. Investigations. After cultures, a baseline blood sample (3 cc) was obtained in all cases. The blood was analysed for a complete blood count and differential on a Coulter counter. Liver function tests and bicarbonate were estimated on an a~toanalyser (Beckman Inc, USA). These investigations were repeated with the 42 hours chloramphenicol levels. Chloramphenicol administration and assay. CS was administered in each dose group in six hourly doses by infusion over 30 minutes. A 3 ml sample of blood was ob-
215
BHU'ITA El" AL : CI ILORAMPHENICOL CLEARANCE WITH TYPHOID FEVER
tained just prior to the 8th dose i.e. at 42 hours for trough estimation. A 3 ml blood sample was drawn one hour after the end of infusion for estimation of peak chloramphenicol levels. Plasma was frozen at -80~ until analysis in fortnightly batches. Analysis was done, on 50 microlitres sample of plasma by high pressure liquid chromatography (HPLC) using the method of Neilsson Ehle et al, 19 with some modifications. (Novapak : microbondapak C 18 column, 25 cm * 2.1 mm id, Waters Associates, USA) was used. Mobile phase was acetic acid, water and methanol in ratios of 1 : 62 : 37. How rate of mobile phase was 1.0 ml per minute. The HPLC system (Waters Associates, USA) consisted of a pump M-(~30 a, a variable LC spectrophotometer 440 ~th absorbance set at 280 mm, WISP 710 B injector, Data Module 730 and a system controller 720. All solvents used were of ultra high purification HPLC grade. Free chloramphenicol for standards was obtained from Parke Davis (Pakistan) Ltd. Standard curve of chloramphenicol was constructed using 5,10,20,30 and 50 meg free chloramphenicol per ml. A four point standard curve, prepared in water, included c0ncentratios of 6.25 to 50 mcg/ml was made. Linear regression analysis of peak heights and chloramphenicol concentrati0ns gave a correlation coefficient of 0.99. Concentration of chloramphenicol in plasma was calculated from the standard curveand reported as mcg chloramphenlcol per ml. Since the analysis was done in fortnightly batches, no therapeutic alteration was made inany group on the basis of the levels estimated. Evaluation of chloramphenicol kinetics.
The chloramphenicol blood levels were fitted to an open single compartment model using an approximation that a steady state had been achieved before the eighth dose after 6-7 half lives had elapsed. The clearance was also calculated on the basis of the area under the curve (AUC) approximation. 2]'22 At steady state the total AUC is proportional to the clearance of the total dose given during the dosing interval, CL = Dose/AUC The AUC was calculated by A U C = ((Cpk + C48)/2)T Where T is the dosing interval, 6 hours. C48 wastaken as C42 or trough level (Figure 1). The basis for these calculations is expressed in Figure 1 and is similar to the approach used by Sack et al.21 Analysis of data was done using an integrated statistical and graphic package, Statgraphics (Statistical Graphics Corporation, USA). The t test was employed for comparison. RESULTS A total of 40 febrile children, admitted to the paediatric ward with suspected typhoid, were placed on CS. Of these 30 were subse-
m
mL~r~
mm J~
~ mB~
~m
~ K ~
'gl
II
4
i
It
i2
"
30
3e
~
48
' n n w In h m n .
Fig. 1.Calculation of pharmacokinetics parameters for steady state assumptions (see text for details)
216
TIlE INDIANJOURaNALOF PEDIATRICS
quently proven to have typhoid on cultures ~nd an alternative diganosis was made in he other ten. The various final diganosis made in the non-typhoid group were viral upper respiratory infection (4), bronchopneumonia (2), otitis media (2), varieella (1) and urinary tract infection (1). The characteristics of the patients in both groups are given in Table 1. The 50 mg/kg/day dosage of CS was stopped after the mid term analysis of the results of the study, as there were two relapses within three weeks of completion of a fourteen day course of therapy on this dosage, and because of concern about the observed low blood levels of chloramphenicol. Figure 2 shows the comparative peak and trough chloramphenicol concentrations for three dosage schedules in 30 children with typhoid. There was no significant difference in the peak (15.4 + 6.1 versus 19.9 _+ 12.2 mcg/ml, mean +_ SD) and trough (5.4 ~- 2.6 versus 8.8 • 7.7 mcg/ml) values on 75 and 100 mg/kg/day C,'S respectively. Three children on 100 mg/kg/day CS had peak chloramphenicol levels in excess of 309 mcg/ml. The mean trough and peak Table 1. Characteristics of the Patient (mean 95% CI) Typhoid N = 30
Non Typhoid n = 10
M.F. 21.9 7.3 Age (yr) 8.6 • 4.0 8.3 • 3.9 Wt (kg) 273 +_ 14.0 23.5 • 8.9 Hepatomegaly (%) 57% 10% Alanine transaminase 64.4 _+33.8 32.9 _+ 16.7 Ou/l) Alkaline phosphate 165.8 +- 53.5 172.1 -+ 66.5 Bicarbonate (mEq/1) 18.9 +_2.4 19.9 _+4.0
Vol. 59, No. 2
"'/'.'....")))7"?)))7')?Z'?)~'?/')I'Z'?)).'" l
A ~
B ~mOtk~d~
C ~00 nOmoSd~r
c~mplm~1~ Do~g, Fig. 2.Compa~son of peak chloramphenico! levels in three dosage groups in patients with
typhoid (Box-whisker plots). Data presented as median (horizontal bar) upper and lower quartiles (boxes) and range. Outliers represented as individal dots. levels of chloramphenicol on 50 mg/kg/day CS in children with typhoid were 4.3 _+ 2.5 and 8.5 • 1.1 mcg/ml respectively. These values were significantly lower than the peak (19.9 _+ i2.2 mcg/ml, p
