581 Four of the strains, shown in the table, had enzyme activities resembling those reported by Brown et al. We did not find
administered for 5 days. to improve. Bumetanide
acetyltransferase, phosphotransferase, or adenylyltransferase activity in the fifth strain examined, although it was highly resistant to gentamicin (minimum inhibitory concentration >128 mg/l). Amikacin and neomycin were not substrates of the phosphotransferase. Kanamycin and tobramycin were mostly poorer substrates than was gentamicin. The chemical structure of gentamicin phosphate has not yet been determined as far as we know, However, an adenylyltransferase that adenylylates the 2"-OH group of gentamicin, kanamycin, and tobramycin is found in some gram-negative organisms.2 We suspect that the phosphotransferase in Staph. aureus attacks the same site and thus is aminoglycoside 2"-O-phosphotransferase. The acetyltransferase is aminoglycoside 6’-N-acetyltransferase since gentamicin CI is not a substrate.3 In contrast to Brown et al. we found acetylation of neomycin. However, acetylation of neomycin and amikacin was poor, and in this the enzyme differed from the aminoglycoside 6’-N-acetyltrans3
day during all events.
Despite this, renal function continued (’Burinex’) 4 mg/day was given every
We think that this case shows that cephalothin sodium and gentamicin in therapeutic doses were well tolerated when given separately whereas in combination they were nephrotoxic. Medical
Department P, Rigshospitalet,
Copenhagen, Denmark
E. TVEDEGAARD
INTERFERENCE WITH GENTAMICIN ASSAY BY HEAT-LABILE SERUM FACTOR
SIR,-Like Dr Shanson and his colleagues,’ we too have observed interference with the microbiological assay of gentamicin by a heat-labile serum factor. Using a four-hour method with Enterobacter cloacce as the assay organism on BHI agar, much higher gentamicin levels were observed in a few clinical
ferase found in some strains of Escherichia coli.2 Department of Microbiology, St Thomas’s Hospital Medical School, London SE1 7EM
K. P. SHANNON IAN PHILLIPS
INTERACTION BETWEEN GENTAMICIN AND CEPHALOTHIN AS CAUSE OP ACUTE RENAL FAILURE
SIR,—Cephalosporins alone or in combination with diuretics,45 gentamicin alone,6 and the combination of cephalosporins and gentamicin have been described as nephrotoxic, the doses generally being rather high.78 We have seen a case suggesting that in combination cephalosporins and gentamicin may be nephrotoxic even in therapeutic doses. A 33-year-old female with no renal disease and a normal plasma-creatinine on admission was treated with methicillin for a Staphylococcus aureus sepsis. Disseminated intravascular coagulation and renal failure developed, and peritoneal dialysis was started. 5 days later a mitral-valve prosthesis had to be inserted because of mitral insufficiency caused by bacterial endocarditis. Methicillin, penicillin, and gentamicin were continued postoperatively for 10 days, and, despite the gentamicin treatment, renal function recovered although proteinuria 1-2 g/day persisted. 23 days after the first operation the prosthesis was replaced because of leak. Methicillin, penicillin, and gentamicin were again administered, but because of a rash all were replaced by cephalothin sodium 50 mg/kg/day, given for 14 days with no effect on renal function, and then withdrawn. 4 days later, because of continuous fever, treatment was started with cephalothin sodium 100 mg/kg/day and gentamicin adjusted after plasma concentrations, which never exceeded 10 µg/ml. After 5 days of treatment the creatinine clearance fell from 50 to 8 mvmin. The antibiotics were withdrawn because no other for the renal failure could be identified despite the comsituation. After that the clearance improved clinical plex slowly, and 1 week later cephalothin sodium was again
cause
Comparison of the microbiological assay using heat-inactivated sera with the radioimmunoassay for gentamicin.
specimens than would be expected for
the dose given and for the time of the blood collection. Since normal sera may have complement-dependent antibacterial activity, we inactivated the sera at 56C before assay and obtained lower values. Now, we routinely inactivate all sera by heating and use Mueller Hinton agar for the Enterobacter assay of gentamicin. Values obtained in this way yield a good linear regression with a radioimmunoassay (Diagnostic Products Corporation), with no significant difference between the two methods with either normal (see figure) or uraemic sera. Departments of Microbiology and Clinical Chemistry, St. Joseph’s Hospital and McMaster University, Hamilton, Ontario, Canada
AUTOMATED TEST FOR BACTERIURIA
2. Price, K. E., Godfrey, J. C., Kawaguchi, H. Adv. appl. Microbiol. 1974, 18, 191.
3. Benveniste, R., Davies, J. Biochemistry, 1971, 10, 1787. 4. Burton, J. R., Lichtenstein, N. S., Colvin, R. B., Hyslop, N. E. Jr. J. Am.
1974, 229, 679. H., Macadam, R. F., Singh, H., Gavras, H., Hartz, S., Turnbull, D., Linton, A. L. J. infect. Dis. 1972, 126, 593. 6. Milman, N. Acta med. scand. 1974, 196, 87. 7 Fillastre, J. P., Laumonier, R., Humbert, G., Dubois, D., Metayer, J., Delpech, A, Leroy, J., Robert, M. Br. med. J. 1973, ii, 396. 8. Opitz, A, Herrmann, I., von Herrath, D., Schaefer, K. Med. Welt. 1971, 22, med. Ass. D.
5. Lawson,
434.
C. ADENIYI-JONES D. STEVENS B. PAGE S. BARBARDORO
test which can distinguish rapidly between speciwhich need detailed microbiological examination and those which do not is of great interest and potential value, but whether full automation of such a test is advantageous is another matter. Dr Johnston and his colleagues (Aug. 21, p. 400) state that the equipment they use is available in most clinical-pathology
SIR,—A
mens
1.
Shanson, D. C., Kensit, J., Hince, C. Lancet, 1975, ii, 875.
582
departments, but it is likely to be in use by chemical pathologists or haematologists, and microbiologists will either have to purchase their own machine or arrange their work to fit in with their colleagues, not an easy thing to do in practice. Most patients with suspected urinary-tract infection attend as outpatients, and if the clinician is not to prescribe unnecessary antibiotic he will have to receive the result before the end
of the clinic when his letter to the patient’s doctor will be dictated, therefore the specimens cannot be batched and dealt with when all have been collected. The great value of automation in chemical pathology is not that it enables a large work load to be handled, but that the results of machine tests are much more accurate. People tire during the day; machines do not. Microbiological specimens have very large innate variation, so this aspect of automation does not apply; moreover, delay in transport will seriously affect the results for most specimens. A laboratory which has a machine capable of dealing with a very large work load will be looking around for specimens to test to justify the cost. Mechanically aided simple tests which can be done as the specimens arrive to recognise the small number of them worthy of detailed examination are what we require. One cannot expect automation developed to solve the problems of a discipline which is concerned with monitoring deviations from normal values to be applicable to our subject where the aim of the investigations is totally different. We need mechanical aid, but small is beautiful in bacteriology. Department of Clinical Pathology, University College Hospital,
E. J. STOKES
London WC1E 6AU
PASTEURELLA AND PETS
SIR,—MR Russell’s letter (Aug. 28, p. 469) on dog bites and Pasteurella prompts us to relate another case of Pasteurella infection which might have had a fatal outcome. Our patient, a man of 54, was admitted for surgery for lung cancer. P. multocida was isolated from a bronchial swab taken at the time of pneumonectomy and from postoperative sputum. It cleared after 5 days’ treatment with co-trimoxazole, and he left hospital 11 days after operation but had to be readmitted 12 days later with an infected pneumonectomy space from which the same organism was isolated. It was suspected that this patient might have had contact with an animal pet and on inquiry it was found that he kept fifteen cats. On hearing of the association with his cats his wife consulted a veterinary surgeon who advised her to get rid of them. The patient’s infection responded satisfactorily to ampicillin and a cephalosporin given systemically and into the pneumonectomy space; he has remained free from infection for 8 weeks. Human infection with Pasteurella can probably be acquired by other means than by biting and may be more common than clinicians suspect. Thoracic and Cardiac Surgical Unit, Harefield Hospital, Harefield, Uxbridge, Middlesex UB9 6JH
JOHN W. JACKSON L. S. NAKHLA
QUANTITATIVE THROAT-SWAB CULTURE SIR,—DR Bell and Dr Smith (July 10, p. 61) conclude that bacteriological examination of throat-swabs for Streptococcus pyogenes is only of diagnostic value when quantitatively assessed, but do not mention the influence of technique in obtaining the specimen. We wish to present the results of studies done to assess the adequacy of specimen taking. swabs taken for viral studies were washed in virus transport medium and the epithelial cells counted in a Neubauer chamber. 38 swabs taken by experienced laboratory staff and 36 by health visitors (State registered nurses) were studied. Cell-counts in those taken by laboratory staff ranged from 22 500 to 667 500 with a median
Throat
of 75 000 cells. per swab. In contrast cell-counts from swabs obtained by health visitors ranged from nil to 300 000 cells with a median count of 22 500 cells per swab. Of these 50% yielded less epithelial cells than the lowest count obtained by laboratory staff, and 8% revealed no epithelial cells. The need to obtain adequate specimens was extensively reported by Ross,’ who also showed that streptococci were not uniformly distributed in the throat. The control samples in the report by Dr Bell and Dr Smith were taken by nursing staff whereas those from patients with pharyngitis were obtained by doctors. We feel that the case for quantitative throat-swab culture cannot be proven unless samples from both groups are obtained by the same personnel. Education of staff in the technique of collection is needed to eliminate what at present amounts to "a wave in the breeze" by some. count
Department of Microbiology, Birmingham Children’s Hospital, Birmingham B16 8ET
R. H. GEORGE P. A. PURDHAM
PRESSURE IN SENGSTAKEN TUBES
SIR,--The guidelines of Baxter et al. 2 for the inflation presof Sengstaken tubes seem to contain a technical error with potentially dangerous consequences. Water-filled non-perfused
sure
manometric catheters fail to reflect true pressures within the oesophagus,3by virtue of sealing of the catheter when a certain pressure is reached. That such an artefact occurred in the study of Baxter et al. is suggested by the fact that, during progressive deflation of the oesophageal balloon, the sensor tube pressure and the balloon pressure coincided when a figure of 20 mm Hg was reached. Until this work is repeated with a more reliable method, physicians using Sengstaken tubes in patients with variceal haemorrhage would be well advised to use inflation pressures which have been proven over the years to stop haemorrhage without causing oesophageal perforation. Gastroenterology Unit, St Vincent’s Hospital, Fitzroy 3065, Victoria, Australia
K. J. BREEN
*,*This letter was shown to the Charing Cross workers, whose reply follows.-ED.L.
SIR,—DR Breen makes a valid technical point about the difference between perfused and non-perfused catheters in the measurement of lower-oesophageal-sphincter (L.E.S.) pressure We are aware of the work of Winans and Harris, but they were recording L.E.S. pressure by pulling the catheter through from stomach to cesophagus. We measured the change in pressure at a fixed point by two rather different methods. The strain-gauge technique is potentially open to error by blockage of the recording orifice, but the catheter was flushed immediately before recordings were taken and the tip had end and side orifices close to each other. We know the mucosa was touching the catheter-tip, but if both orifices were plugged by mucus or adherent mucosa there would have been little drop in pressure with deflation of the balloon, and it would not have reached zero. We also measured pressures during progressive inflation of the balloon and obtained a mirror image of the curves shown in our figure. The much simpler method we advocated at the end of the article measured the "opening" pressure-in other words, the pressure required to separate the balloon from the oesophageai wall around the tip of the catheter. If the syringe is elevated until water flows and then lowered until it stops flowing, the potential hazard of a blocked catheter is overcome. We now use a standard central-venous-pressure line attached to the
thoroughly
1. 2.
Ross, P. W. Practitioner, 1971, 207, 791. Baxter, H. K., Kirk, C. J. C., Johnson, A. G., Murray-Lyon, Reynolds, K. W. Lancet, 1976, i, 1053. 3. Winans, C. S., Harris, L. D. Gastroenterology, 1967, 52, 773.
I. M.,
administered for 5 days. to improve. Bumetanide
acetyltransferase, phosphotransferase, or adenylyltransferase activity in the fifth strain examined, although it was highly resistant to gentamicin (minimum inhibitory concentration >128 mg/l). Amikacin and neomycin were not substrates of the phosphotransferase. Kanamycin and tobramycin were mostly poorer substrates than was gentamicin. The chemical structure of gentamicin phosphate has not yet been determined as far as we know, However, an adenylyltransferase that adenylylates the 2"-OH group of gentamicin, kanamycin, and tobramycin is found in some gram-negative organisms.2 We suspect that the phosphotransferase in Staph. aureus attacks the same site and thus is aminoglycoside 2"-O-phosphotransferase. The acetyltransferase is aminoglycoside 6’-N-acetyltransferase since gentamicin CI is not a substrate.3 In contrast to Brown et al. we found acetylation of neomycin. However, acetylation of neomycin and amikacin was poor, and in this the enzyme differed from the aminoglycoside 6’-N-acetyltrans3
day during all events.
Despite this, renal function continued (’Burinex’) 4 mg/day was given every
We think that this case shows that cephalothin sodium and gentamicin in therapeutic doses were well tolerated when given separately whereas in combination they were nephrotoxic. Medical
Department P, Rigshospitalet,
Copenhagen, Denmark
E. TVEDEGAARD
INTERFERENCE WITH GENTAMICIN ASSAY BY HEAT-LABILE SERUM FACTOR
SIR,-Like Dr Shanson and his colleagues,’ we too have observed interference with the microbiological assay of gentamicin by a heat-labile serum factor. Using a four-hour method with Enterobacter cloacce as the assay organism on BHI agar, much higher gentamicin levels were observed in a few clinical
ferase found in some strains of Escherichia coli.2 Department of Microbiology, St Thomas’s Hospital Medical School, London SE1 7EM
K. P. SHANNON IAN PHILLIPS
INTERACTION BETWEEN GENTAMICIN AND CEPHALOTHIN AS CAUSE OP ACUTE RENAL FAILURE
SIR,—Cephalosporins alone or in combination with diuretics,45 gentamicin alone,6 and the combination of cephalosporins and gentamicin have been described as nephrotoxic, the doses generally being rather high.78 We have seen a case suggesting that in combination cephalosporins and gentamicin may be nephrotoxic even in therapeutic doses. A 33-year-old female with no renal disease and a normal plasma-creatinine on admission was treated with methicillin for a Staphylococcus aureus sepsis. Disseminated intravascular coagulation and renal failure developed, and peritoneal dialysis was started. 5 days later a mitral-valve prosthesis had to be inserted because of mitral insufficiency caused by bacterial endocarditis. Methicillin, penicillin, and gentamicin were continued postoperatively for 10 days, and, despite the gentamicin treatment, renal function recovered although proteinuria 1-2 g/day persisted. 23 days after the first operation the prosthesis was replaced because of leak. Methicillin, penicillin, and gentamicin were again administered, but because of a rash all were replaced by cephalothin sodium 50 mg/kg/day, given for 14 days with no effect on renal function, and then withdrawn. 4 days later, because of continuous fever, treatment was started with cephalothin sodium 100 mg/kg/day and gentamicin adjusted after plasma concentrations, which never exceeded 10 µg/ml. After 5 days of treatment the creatinine clearance fell from 50 to 8 mvmin. The antibiotics were withdrawn because no other for the renal failure could be identified despite the comsituation. After that the clearance improved clinical plex slowly, and 1 week later cephalothin sodium was again
cause
Comparison of the microbiological assay using heat-inactivated sera with the radioimmunoassay for gentamicin.
specimens than would be expected for
the dose given and for the time of the blood collection. Since normal sera may have complement-dependent antibacterial activity, we inactivated the sera at 56C before assay and obtained lower values. Now, we routinely inactivate all sera by heating and use Mueller Hinton agar for the Enterobacter assay of gentamicin. Values obtained in this way yield a good linear regression with a radioimmunoassay (Diagnostic Products Corporation), with no significant difference between the two methods with either normal (see figure) or uraemic sera. Departments of Microbiology and Clinical Chemistry, St. Joseph’s Hospital and McMaster University, Hamilton, Ontario, Canada
AUTOMATED TEST FOR BACTERIURIA
2. Price, K. E., Godfrey, J. C., Kawaguchi, H. Adv. appl. Microbiol. 1974, 18, 191.
3. Benveniste, R., Davies, J. Biochemistry, 1971, 10, 1787. 4. Burton, J. R., Lichtenstein, N. S., Colvin, R. B., Hyslop, N. E. Jr. J. Am.
1974, 229, 679. H., Macadam, R. F., Singh, H., Gavras, H., Hartz, S., Turnbull, D., Linton, A. L. J. infect. Dis. 1972, 126, 593. 6. Milman, N. Acta med. scand. 1974, 196, 87. 7 Fillastre, J. P., Laumonier, R., Humbert, G., Dubois, D., Metayer, J., Delpech, A, Leroy, J., Robert, M. Br. med. J. 1973, ii, 396. 8. Opitz, A, Herrmann, I., von Herrath, D., Schaefer, K. Med. Welt. 1971, 22, med. Ass. D.
5. Lawson,
434.
C. ADENIYI-JONES D. STEVENS B. PAGE S. BARBARDORO
test which can distinguish rapidly between speciwhich need detailed microbiological examination and those which do not is of great interest and potential value, but whether full automation of such a test is advantageous is another matter. Dr Johnston and his colleagues (Aug. 21, p. 400) state that the equipment they use is available in most clinical-pathology
SIR,—A
mens
1.
Shanson, D. C., Kensit, J., Hince, C. Lancet, 1975, ii, 875.
582
departments, but it is likely to be in use by chemical pathologists or haematologists, and microbiologists will either have to purchase their own machine or arrange their work to fit in with their colleagues, not an easy thing to do in practice. Most patients with suspected urinary-tract infection attend as outpatients, and if the clinician is not to prescribe unnecessary antibiotic he will have to receive the result before the end
of the clinic when his letter to the patient’s doctor will be dictated, therefore the specimens cannot be batched and dealt with when all have been collected. The great value of automation in chemical pathology is not that it enables a large work load to be handled, but that the results of machine tests are much more accurate. People tire during the day; machines do not. Microbiological specimens have very large innate variation, so this aspect of automation does not apply; moreover, delay in transport will seriously affect the results for most specimens. A laboratory which has a machine capable of dealing with a very large work load will be looking around for specimens to test to justify the cost. Mechanically aided simple tests which can be done as the specimens arrive to recognise the small number of them worthy of detailed examination are what we require. One cannot expect automation developed to solve the problems of a discipline which is concerned with monitoring deviations from normal values to be applicable to our subject where the aim of the investigations is totally different. We need mechanical aid, but small is beautiful in bacteriology. Department of Clinical Pathology, University College Hospital,
E. J. STOKES
London WC1E 6AU
PASTEURELLA AND PETS
SIR,—MR Russell’s letter (Aug. 28, p. 469) on dog bites and Pasteurella prompts us to relate another case of Pasteurella infection which might have had a fatal outcome. Our patient, a man of 54, was admitted for surgery for lung cancer. P. multocida was isolated from a bronchial swab taken at the time of pneumonectomy and from postoperative sputum. It cleared after 5 days’ treatment with co-trimoxazole, and he left hospital 11 days after operation but had to be readmitted 12 days later with an infected pneumonectomy space from which the same organism was isolated. It was suspected that this patient might have had contact with an animal pet and on inquiry it was found that he kept fifteen cats. On hearing of the association with his cats his wife consulted a veterinary surgeon who advised her to get rid of them. The patient’s infection responded satisfactorily to ampicillin and a cephalosporin given systemically and into the pneumonectomy space; he has remained free from infection for 8 weeks. Human infection with Pasteurella can probably be acquired by other means than by biting and may be more common than clinicians suspect. Thoracic and Cardiac Surgical Unit, Harefield Hospital, Harefield, Uxbridge, Middlesex UB9 6JH
JOHN W. JACKSON L. S. NAKHLA
QUANTITATIVE THROAT-SWAB CULTURE SIR,—DR Bell and Dr Smith (July 10, p. 61) conclude that bacteriological examination of throat-swabs for Streptococcus pyogenes is only of diagnostic value when quantitatively assessed, but do not mention the influence of technique in obtaining the specimen. We wish to present the results of studies done to assess the adequacy of specimen taking. swabs taken for viral studies were washed in virus transport medium and the epithelial cells counted in a Neubauer chamber. 38 swabs taken by experienced laboratory staff and 36 by health visitors (State registered nurses) were studied. Cell-counts in those taken by laboratory staff ranged from 22 500 to 667 500 with a median
Throat
of 75 000 cells. per swab. In contrast cell-counts from swabs obtained by health visitors ranged from nil to 300 000 cells with a median count of 22 500 cells per swab. Of these 50% yielded less epithelial cells than the lowest count obtained by laboratory staff, and 8% revealed no epithelial cells. The need to obtain adequate specimens was extensively reported by Ross,’ who also showed that streptococci were not uniformly distributed in the throat. The control samples in the report by Dr Bell and Dr Smith were taken by nursing staff whereas those from patients with pharyngitis were obtained by doctors. We feel that the case for quantitative throat-swab culture cannot be proven unless samples from both groups are obtained by the same personnel. Education of staff in the technique of collection is needed to eliminate what at present amounts to "a wave in the breeze" by some. count
Department of Microbiology, Birmingham Children’s Hospital, Birmingham B16 8ET
R. H. GEORGE P. A. PURDHAM
PRESSURE IN SENGSTAKEN TUBES
SIR,--The guidelines of Baxter et al. 2 for the inflation presof Sengstaken tubes seem to contain a technical error with potentially dangerous consequences. Water-filled non-perfused
sure
manometric catheters fail to reflect true pressures within the oesophagus,3by virtue of sealing of the catheter when a certain pressure is reached. That such an artefact occurred in the study of Baxter et al. is suggested by the fact that, during progressive deflation of the oesophageal balloon, the sensor tube pressure and the balloon pressure coincided when a figure of 20 mm Hg was reached. Until this work is repeated with a more reliable method, physicians using Sengstaken tubes in patients with variceal haemorrhage would be well advised to use inflation pressures which have been proven over the years to stop haemorrhage without causing oesophageal perforation. Gastroenterology Unit, St Vincent’s Hospital, Fitzroy 3065, Victoria, Australia
K. J. BREEN
*,*This letter was shown to the Charing Cross workers, whose reply follows.-ED.L.
SIR,—DR Breen makes a valid technical point about the difference between perfused and non-perfused catheters in the measurement of lower-oesophageal-sphincter (L.E.S.) pressure We are aware of the work of Winans and Harris, but they were recording L.E.S. pressure by pulling the catheter through from stomach to cesophagus. We measured the change in pressure at a fixed point by two rather different methods. The strain-gauge technique is potentially open to error by blockage of the recording orifice, but the catheter was flushed immediately before recordings were taken and the tip had end and side orifices close to each other. We know the mucosa was touching the catheter-tip, but if both orifices were plugged by mucus or adherent mucosa there would have been little drop in pressure with deflation of the balloon, and it would not have reached zero. We also measured pressures during progressive inflation of the balloon and obtained a mirror image of the curves shown in our figure. The much simpler method we advocated at the end of the article measured the "opening" pressure-in other words, the pressure required to separate the balloon from the oesophageai wall around the tip of the catheter. If the syringe is elevated until water flows and then lowered until it stops flowing, the potential hazard of a blocked catheter is overcome. We now use a standard central-venous-pressure line attached to the
thoroughly
1. 2.
Ross, P. W. Practitioner, 1971, 207, 791. Baxter, H. K., Kirk, C. J. C., Johnson, A. G., Murray-Lyon, Reynolds, K. W. Lancet, 1976, i, 1053. 3. Winans, C. S., Harris, L. D. Gastroenterology, 1967, 52, 773.
I. M.,
