437
study from our clinic suggesting that doses of levothyroxine achieving TSH levels of 0-4 mU/1 (lower limit of normal) result in maximum lowering of the serum thyroglobulin levels in patients with metastatic thyroid cancer who have been subjected to thyroid ablation. This finding has discouraged us from attempting more vigorous treatment with levothyroxine designed to reach the lower limits of detectability afforded by fourth-generation TSH assays. Endocrinology and Metabolism, Department of Medicine, University of Minnesota, Minneapolis, Minnesota 55455, USA
Section of
JAMES D. WALKER CARY N. MARIASH
JACK H. OPPENHEIMER
DS, Neer RM, Ridgway EC, Daniels GH. Subclinical hyperthyroidism and reduced bone mineral density as a possible result of prolonged suppression of the pituitary-thyroid axis with L-thyroxine. Am J Med 1987; 82: 1167-70. 2. Paul TL, Kerrigan J, Kelly AM, Braverman LE, Baran DT. Long-term L-thyroxine therapy is associated with decreased hip bone density in premenopausal women. 1. Ross
JAMA 1988, 259: 3137-41. 3. Diamond T, Nery L, Hales I. A therapeutic dilemma: suppressive doses
of thyroxine significantly reduce bone mineral measurements in both premenopausal and women with Clin Endocrinol Metab 1990; 72: postmenopausal thyroid carcinoma. J
1184-88. 4 Burmeister LA, Goumaz MO, Manash CN, Oppenheimer JH. Levothyroxine dose requirements for thyrotropin suppression in the treatment of differentiated thyroid cancer. J Clin Endocrinol Metab (in press).
SIR,- We have investigated the influence of calcitonin deficiency bone turnover and density in 25 premenopausal female and 12 male patients aged 23-49 who had undergone total thyroidectomy for differentiated thyroid cancer 1-15 years previously.l Basal and calcium stimulated extractable calcitonin (monomeric, biologically active form of the hormone) was lacking or much decreased in all patients. There was a relative increase of urine hydroxyproline excretion (an index of osteoclastic bone degradation) in relation to serum osteocalcin (an index of osteoblastic bone formation), indicating an imbalance of bone turnover with a tendency to increased degradation in all patients. Total and trabecular bone density, measured by computed tomography at the distal forearm, were significantly decreased in the male and normal in the female patients unrelated to the duration of the calcitonin deficiency: Bone density on
Women
(mg/ml) Total Patients Controls
p* Trabecular Patients Controls
p*
299 295
(53) (n 25) (52) (n -12) =
Men 294 308
NS
150 (15) (n = 25) 156 (17) (n = 12) NS
(48) (n = 12) (38) (n = 14) 0-019
155 189
(19) (n= 12) (21) (n = 14)
as mean (SD) indicates that patients with calcitonin deficiency studyl and;or suppressive thyroid hormone treatment may be at increased risk of bone degradation and osteopenia. The fact that only the male patients had decreased bone density may be due to lower parathyroid activity in our female patients and the greater thyroidectomy-induced decrement of monomeric calcitonin in male patients below that in male controls. Dr Franklyn and colleagues’ study lacks an investigation of the influence of calcitonin deficiency. BMD values for the 5 male patients, but not for the females, do suggest significantly reduced values. A larger number of males in that study might have confirmed our observations on bone density in thyroidectomised male patients. We still do not know if suppressive doses of thyroxine or calcitonin deficiency can cause osteopenia (at least in men). Franklyn et al used a relatively insensitive method to measure bone density, peripheral or axial quantitative computed tomography is more sensitive to early changes in trabecular bone density.2,3 Follow-up studies with a method of higher diagnostic sensitivity may provide more conclusive information.4
Our
Luitpoldkrankenhaus, Wurzburg. Germany
8700
1. Schneider P,
PETER SCHNEIDER WILHELM BÖRNER
Berger P, Kruse K, Börner W Effect of calcitonin deficiency on bone in rotally thyroidectomized patients.J Endocrinol Invest density and bone turnover in 1991; 14: 935-42
Leukaemia after iodine-131 exposure SIR,-Dr Hall and colleagues’ thorough study of the incidence of leukaemia after iodine-131 exposure (July 4, p 1) revealed no sign of a radiation effect in 47 000 patients. They suggest that the explanation may be a reduction in effect because of the protracted radiation from 8-day half-life iodine-131 compared with that of a single acute exposure from atomic bombs in Japan. However, X-ray treatment given to patients with ankylosing spondylitis was protracted and has never been suggested to cause a lower incidence of leukaemia because of this. Only 3% of the Swedish patients in Hall’s study received bone marrow doses over 100 mGy. Is it heretical to surmise that there may be a threshold dose? The absence of a threshold is an old dogma but Pochin used to remind us that if only a single strand of DNA is broken, the cell’s enzyme systems can and will rapidly and correctly repair the break.’ The Bowies’ doubts about risks from radon exposure2 and Brewin’s attitude to "dilute radiation"3 may be sensible. There are real worries about the effects of nuclear weapons, and an almost atavistic fear of cancer persists. These fears become confused with radiation, which becomes the scapegoat. Some epidemiologists seem to relax their standards when looking for radiation effects from low levels. The study group is often found to have a lower cancer incidence than the general population and some form of selection such as the convenient but imprecise "healthy worker effect" is invoked. Then the cancers are subdivided into many subgroups and one or two are found to have a slight excess, which is reported and accepted, the deficiencies being ignored. Calculations of potential cancer risk from low-level radiation that include unproven effects from the lowest doses cause widespread concern, so it is not surprising that the public perception of radiation is so hostile. 2 Wensleydale Gardens, Hampton, TW12 2LU, UK
KEITH E. HALNAN
0-023
*p by Mann-Whitney U test, data
Nuclear Medicine Clinic, University of Wurzburg,
2. Schneider P, Borner W. Significance of two different bone mineral measurement methods in the assessment of the peripheral and axial skeleton. In: Ring EFJ, ed. Current research in osteoporosis and bone mineral measurement II. London: British Institute of Radiology; 1992: 18. 3. Whitehouse RW, Karantanas A, Adams JE. Discrepancies in spinal bone mass measured by QCT and DXA. In: Ring EFJ, ed. Current research in osteoporosis and bone mineral measurement II. London: British Institute of Radiology, 1992: 26. 4. Reiners C, Doerr K, Rohrbach C, Stoesser D, Jockenhovel F. Vorzeitige osteoporose bei pramenopausalen Frauen unter TSH-suppressiver Levothyroxin-Medikation nach Thyreoidektomie und Radiojodtherapie wegen Schilddrüsenkarzinom? Nucl Med 1991; 30: A69.
1 Pochin EE. Nuclear radiation: risks and benefits. Oxford. Clarendon, 1983: 92-97. 2. Bowie C, Bowie SHU. Radon and health. Lancet 1991; 337: 409-13.
3. Brewin TB. Excessive fear of dilute radiation Proc R Soc Med 1992; 85: 311-13.
Antibiotic resistance in pneumococcal
meningitis SIR,-In 1991 Pecoul and colleagues1 reported their experience chloramphenicol or ampicillin in the management of meningitis in west Africa. The in-hospital case fatality rate was 33% but for pneumococcal meningitis it was 67% (77/115), and 8 of the 38 survivors had serious neurological sequelae-ie, there was an unsatisfactory outcome in 74 % of children. Their conclusion, in the face of a 74% failure rate, that long-acting chloramphenicol is the with
antibiotic of choice as first-line treatment, suggests that constraints of cost now have resulted in there being no effective management for this condition in the third world countries of Africa. Why did ampicillin or chloramphenicol fail so often? Pecoul et al do not analyse the efficacy of the two agents separately, but we have reported the failure of penicillin or chloramphenicol in penicillinresistant pneumococcal meningitis.2 Pecoul et al report 9-1% resistance to chloramphenicol but only 1-5% (1/66) resistance to ampicillin. Resistance was 41-7% in our series. Have their laboratory methods failed to detect penicillin intermediate-resistant strains responsible for considerable morbidity and mortality? Their diffusion susceptibilty method (Pasteur Diagnostics) uses a 10 Irg ampicillin disc. A zone size of 29 mm is used to identify
438
intermediate and 21 mm to identify fully resistant strains. Jacobs et al showed in 19803 that these discs lack sensitivity and specificity, analysis of their histogram suggesting that 897%-966% of intermediately resistant pneumococci would be missed with a 10 µg ampicillin disc at a cut-off of 21 mm and at least 37-9% at 29 mm. If Pecoul et al used the former cut-off the great majority of intermediately resistant strains might not have been identified as such. A 1 µg oxacillin disk is the preferred method of
Because of the previous rarity of PRP, it has been usual practice in the UK to treat patients with proven pneumococcal meningitis with high doses of benzylpenicillin intravenously without waiting for antibiotic sensitivity results. It would now be preferable to initiate treatment with a third-generation cephalosporin such as cefotaxime. Therapy can be changed to penicillin if the strain is subsequently shown to be sensitive by appropriate testing.
evaluating pneumococcal susceptibiliry4 to penicillin/ampicillin. Chloramphenicol-resistant pneumococci were first isolated in west
Department of Medical Microbiology, Guy’s Hospital, London SE1 9RT, UK
Africa in 1978 but since that time few data, obtained with proper methods, are available on the prevalence of penicillin or ampicillinresistant strains in sub-Saharan Africa, north of South Africa. The data of Pecoul et al argue the need for information on the prevalence of penicillin-resistant pneumococcal strains in third world countries. Unfortunately, the third-generation cephalosporins required to treat these infections are unavailable in these countries on the grounds of cost. Department of Medical Microbiology, MRC Emergent Pathogen Research Unit, South African Institute for Medical Research, and University of the Witwatersrand, PO Box 1038, Johannesburg 2000, South Africa
KEITH P. KLUGMAN HENDRICK J. KOORNHOF
Department of Paediatrics, Baragwanath Hospital, Soweto
IAN R. FRIEDLAND
F, Keita M, et al. Long-acting chloramphenicol versus intravenous ampicillin for treatment of bacterial mengingitis. Lancet 1991, 338: 862-65. 2. Friedland IR, Klugman KP. Failure of chloramphenicol therapy in penicillin-resistant pneumococcal meningitis. Lancet 1992; 339: 405-08. 3. Jacobs MR, Gasper MN, Robins-Browne RM, Koornhof HJ. Antimicrobial susceptibility testing of pneumococci II. determination of optimal disc diffusion test for detection of penicillin resistance. J Antimicrob Chemother 1980; 6: 53-64. 4. Klugman KP. Pneumococcal resistance to antibiotics. Clin Microbiol Rev 1990; 3: 1. Pécoul B, Varaine
171-96.
SIR,- The prevalence of penicillin-resistant Streptococcus pneumoniae (PRP) varies greatly, resistance rates of 0% in France,’ 24% in South Africa,2 and 38% in Spain3having been reported. In the UK, PRP is uncommon but, as Ridgeway and colleagues have pointed out’ the rate is increasing and there was an almost four-fold rise in the number of PRP referred to the Central Public Health Laboratory from 1987 to 1991.5 We report here another case of PRP meningitis in which the organism was initially reported as penicillin sensitive. A 72-year-old woman was admitted to a London hospital in January, 1992, with symptoms of meningitis and a fever of 38.5°C. Patchy consolidation of the right lung field was seen on chest X-ray. The white blood cell count was 35 x 109/1 (92% neutrophils). A lumbar puncture yielded turbid cerebrospinal fluid (CSF) (white cell count 2-4 x 109/1 [95% neutrophils], protein 3-8 g/1, and glucose 01 rnmol/1 [peripheral blood glucose 7-4 mmol/1]). Gram-positive diplococci were seen on a CSF smear. The patient was admitted to intensive care, sedated, and intubated. She was put on intravenous ceftazidime 2 g every 6 h. S pneumoniae was grown from the CSF and was reported as sensitive to penicillin (1 µg penicillin disc), ampicillin, and ceftazidime. The organism was also grown from blood cultures. On day 2, the patient’s renal function appeared to be deteriorating and she was transferred to Guy’s Hospital. Antibiotic therapy was changed to benzylpenicillin 2 megaunits every 4 h. Pneumococci were again grown from blood cultures but on sensitivity testing, there was no zone around a 1 µg oxacillin disc. A computed tomographic scan of the brain revealed no abnormality. The patient recovered slowly and was discharged from intensive care 2 weeks later. The organism was a type 9 pneumococcus which had penicillin minimum inhibitory and bactericidal concentrations of 1 mg/1. This case illustrates the importance of testing isolates of S pneumoniae against a 1 Ilg oxacillin disc if penicillin resistance is to be detected. A study in 1987 showed that laboratories using oxacillin discs made fewer errors in detecting resistance or moderate resistance of pneumococci to penicillin than those using high content penicillin discs alone.6 This is especially important in meningitis, for which penicillin is inadequate as treatment even with moderately resistant strains (MIC 0’1-1 mg/1).
M. I. ISSACK N. A. SIMMONS G. L. FRENCH
AY, Goldstein FW, Acar JF. A seventeen year epidemiological survey of antimicrobial resistance in pneumococci in two hospitals. J Antimicrob Chemother 1988; 22 (suppl B). 41-52. 2 Klugman KP, Koornhof HJ, Gasas A, Storey K, Gilbertson I Carnage of penicillin-resistant pneumococci. Arch Dis Child 1986; 61: 377-81. 3. Casal J, Fenoll A, Vicioso M, Munoz R. Increase in resistance to penicillin in pneumococci in Spain. Lancet 1989; i: 735. 4. Ridgway EJ, Allen KD, Neal TJ, Lombard M, Rigby A Penicillin-resistant pneumococcal meningitis. Lancet 1992; 339: 931. 5. George RC, Ball LC, Cooper PG. Antibiotic resistant pneumococci in the United Kingdom. CDR Rev 1992; 2: 37-13. 6. Snell JJS, George RC, Perry SF, Erdman YJ. Antimicrobial susceptibility testing of Streptococcus pneumomae quality assessment results. J Clin Pathol 1988, 41: 1. Buu-Hoi
384-87.
SIR,-Recent communications on chloramphenicol in bactenal meningitis (predominantly pneumococcal) may cause anxiety among clinicians in the developing world, where the therapeutic options are already limited. We are often fortunate if we can identify the causative organism causing infection, let alone its antibiotic sensitivity, and the choice of antibiotic is often a matter of expediency rather than one based on hard micobiological data. Friedland and Klugman’s prospective studyl and the casereports of George et aP and Dr Corcoran and colleagues (July 4, p 50) describe initial treatments of up to 48 h with penicillin in combination with chloramphenicol and resultant poor responses in patients with penicillin-resistant pneumococcal meningitis. Penicillin was used in two instances even when resistance had been noted. It would have been helpful (especially in the prospective study) to know, in the presence of relative penicillin insensitivity, whether there was antagonism as measured by fractional lethal concentration indexes. In a wide range of organisms antagonism between penicillins and chloramphenicol has been described in which chloramphenicol itself was acting only as a bacteriostatic agent.3,4 The 48 % of children in the penicillin-resistant group who required alteration of antibiotic from chloramphenicol’ might represent a subgroup of patients who were relatively untreated during the first 48 h. Where bacterial meningitis has to be treated "blind" few clinicians with access to third-generation cephalosporins would not use them. Chloramphenicol alone in bacterial meningitis has a good record in developing countries5,6 and if penicillin resistance in
pneumococci were automatically to preclude chloramphenicol use, many adults and children would be in effect untreatable in many countries. The interaction of penicillin with chloramphenicol in
penicillin-insensitive organisms and the role of chloramphenicol as primary monotherapy need to be more clearly defined before there is a move away from chloramphenicol in the treatment of bacterial meningitis in developing countries. Department of Paediatrics, Royal Victoria Hospital, Banjul, The Gambia
THOMAS MCKAY
1. Friedland IR, Klugman KP Failure of 2.
chloramphenicol therapy in penicillin-resistant pneumococcal meningitis Lancet 1992; 339: 405-08 George RE, Wariyar UK, Nao GG. Penicillin-resistant pneumococcal meningitis
Lancet 1992; 339: 931-32. 3 Mathies AW, Leedom JM, Ivler D, et al. Antibiotic antagonism in bacterial meningitis. Anitmicrob Agents Chemother 1967, 7: 218-24. 4. Weeks JL, Mason EO, Baker CJ Antagonism of ampicillin and chloramphenicol for meningeal isolates of group B streptococci Antimicrob Agents Chemother 1981, 20: 281-85. 5. Shann F, Barker J, Poore P. Chloramphenicol alone versus chloramphenicol plus penicillin for bacterial meningitis m children Lancet 1985, ii: 681-84 6 McKay T Experience of changing antibiotic protocol in childhood bacterial meningitis in Vanuatu. Trop Doc 1989; 19: 158-59.
study from our clinic suggesting that doses of levothyroxine achieving TSH levels of 0-4 mU/1 (lower limit of normal) result in maximum lowering of the serum thyroglobulin levels in patients with metastatic thyroid cancer who have been subjected to thyroid ablation. This finding has discouraged us from attempting more vigorous treatment with levothyroxine designed to reach the lower limits of detectability afforded by fourth-generation TSH assays. Endocrinology and Metabolism, Department of Medicine, University of Minnesota, Minneapolis, Minnesota 55455, USA
Section of
JAMES D. WALKER CARY N. MARIASH
JACK H. OPPENHEIMER
DS, Neer RM, Ridgway EC, Daniels GH. Subclinical hyperthyroidism and reduced bone mineral density as a possible result of prolonged suppression of the pituitary-thyroid axis with L-thyroxine. Am J Med 1987; 82: 1167-70. 2. Paul TL, Kerrigan J, Kelly AM, Braverman LE, Baran DT. Long-term L-thyroxine therapy is associated with decreased hip bone density in premenopausal women. 1. Ross
JAMA 1988, 259: 3137-41. 3. Diamond T, Nery L, Hales I. A therapeutic dilemma: suppressive doses
of thyroxine significantly reduce bone mineral measurements in both premenopausal and women with Clin Endocrinol Metab 1990; 72: postmenopausal thyroid carcinoma. J
1184-88. 4 Burmeister LA, Goumaz MO, Manash CN, Oppenheimer JH. Levothyroxine dose requirements for thyrotropin suppression in the treatment of differentiated thyroid cancer. J Clin Endocrinol Metab (in press).
SIR,- We have investigated the influence of calcitonin deficiency bone turnover and density in 25 premenopausal female and 12 male patients aged 23-49 who had undergone total thyroidectomy for differentiated thyroid cancer 1-15 years previously.l Basal and calcium stimulated extractable calcitonin (monomeric, biologically active form of the hormone) was lacking or much decreased in all patients. There was a relative increase of urine hydroxyproline excretion (an index of osteoclastic bone degradation) in relation to serum osteocalcin (an index of osteoblastic bone formation), indicating an imbalance of bone turnover with a tendency to increased degradation in all patients. Total and trabecular bone density, measured by computed tomography at the distal forearm, were significantly decreased in the male and normal in the female patients unrelated to the duration of the calcitonin deficiency: Bone density on
Women
(mg/ml) Total Patients Controls
p* Trabecular Patients Controls
p*
299 295
(53) (n 25) (52) (n -12) =
Men 294 308
NS
150 (15) (n = 25) 156 (17) (n = 12) NS
(48) (n = 12) (38) (n = 14) 0-019
155 189
(19) (n= 12) (21) (n = 14)
as mean (SD) indicates that patients with calcitonin deficiency studyl and;or suppressive thyroid hormone treatment may be at increased risk of bone degradation and osteopenia. The fact that only the male patients had decreased bone density may be due to lower parathyroid activity in our female patients and the greater thyroidectomy-induced decrement of monomeric calcitonin in male patients below that in male controls. Dr Franklyn and colleagues’ study lacks an investigation of the influence of calcitonin deficiency. BMD values for the 5 male patients, but not for the females, do suggest significantly reduced values. A larger number of males in that study might have confirmed our observations on bone density in thyroidectomised male patients. We still do not know if suppressive doses of thyroxine or calcitonin deficiency can cause osteopenia (at least in men). Franklyn et al used a relatively insensitive method to measure bone density, peripheral or axial quantitative computed tomography is more sensitive to early changes in trabecular bone density.2,3 Follow-up studies with a method of higher diagnostic sensitivity may provide more conclusive information.4
Our
Luitpoldkrankenhaus, Wurzburg. Germany
8700
1. Schneider P,
PETER SCHNEIDER WILHELM BÖRNER
Berger P, Kruse K, Börner W Effect of calcitonin deficiency on bone in rotally thyroidectomized patients.J Endocrinol Invest density and bone turnover in 1991; 14: 935-42
Leukaemia after iodine-131 exposure SIR,-Dr Hall and colleagues’ thorough study of the incidence of leukaemia after iodine-131 exposure (July 4, p 1) revealed no sign of a radiation effect in 47 000 patients. They suggest that the explanation may be a reduction in effect because of the protracted radiation from 8-day half-life iodine-131 compared with that of a single acute exposure from atomic bombs in Japan. However, X-ray treatment given to patients with ankylosing spondylitis was protracted and has never been suggested to cause a lower incidence of leukaemia because of this. Only 3% of the Swedish patients in Hall’s study received bone marrow doses over 100 mGy. Is it heretical to surmise that there may be a threshold dose? The absence of a threshold is an old dogma but Pochin used to remind us that if only a single strand of DNA is broken, the cell’s enzyme systems can and will rapidly and correctly repair the break.’ The Bowies’ doubts about risks from radon exposure2 and Brewin’s attitude to "dilute radiation"3 may be sensible. There are real worries about the effects of nuclear weapons, and an almost atavistic fear of cancer persists. These fears become confused with radiation, which becomes the scapegoat. Some epidemiologists seem to relax their standards when looking for radiation effects from low levels. The study group is often found to have a lower cancer incidence than the general population and some form of selection such as the convenient but imprecise "healthy worker effect" is invoked. Then the cancers are subdivided into many subgroups and one or two are found to have a slight excess, which is reported and accepted, the deficiencies being ignored. Calculations of potential cancer risk from low-level radiation that include unproven effects from the lowest doses cause widespread concern, so it is not surprising that the public perception of radiation is so hostile. 2 Wensleydale Gardens, Hampton, TW12 2LU, UK
KEITH E. HALNAN
0-023
*p by Mann-Whitney U test, data
Nuclear Medicine Clinic, University of Wurzburg,
2. Schneider P, Borner W. Significance of two different bone mineral measurement methods in the assessment of the peripheral and axial skeleton. In: Ring EFJ, ed. Current research in osteoporosis and bone mineral measurement II. London: British Institute of Radiology; 1992: 18. 3. Whitehouse RW, Karantanas A, Adams JE. Discrepancies in spinal bone mass measured by QCT and DXA. In: Ring EFJ, ed. Current research in osteoporosis and bone mineral measurement II. London: British Institute of Radiology, 1992: 26. 4. Reiners C, Doerr K, Rohrbach C, Stoesser D, Jockenhovel F. Vorzeitige osteoporose bei pramenopausalen Frauen unter TSH-suppressiver Levothyroxin-Medikation nach Thyreoidektomie und Radiojodtherapie wegen Schilddrüsenkarzinom? Nucl Med 1991; 30: A69.
1 Pochin EE. Nuclear radiation: risks and benefits. Oxford. Clarendon, 1983: 92-97. 2. Bowie C, Bowie SHU. Radon and health. Lancet 1991; 337: 409-13.
3. Brewin TB. Excessive fear of dilute radiation Proc R Soc Med 1992; 85: 311-13.
Antibiotic resistance in pneumococcal
meningitis SIR,-In 1991 Pecoul and colleagues1 reported their experience chloramphenicol or ampicillin in the management of meningitis in west Africa. The in-hospital case fatality rate was 33% but for pneumococcal meningitis it was 67% (77/115), and 8 of the 38 survivors had serious neurological sequelae-ie, there was an unsatisfactory outcome in 74 % of children. Their conclusion, in the face of a 74% failure rate, that long-acting chloramphenicol is the with
antibiotic of choice as first-line treatment, suggests that constraints of cost now have resulted in there being no effective management for this condition in the third world countries of Africa. Why did ampicillin or chloramphenicol fail so often? Pecoul et al do not analyse the efficacy of the two agents separately, but we have reported the failure of penicillin or chloramphenicol in penicillinresistant pneumococcal meningitis.2 Pecoul et al report 9-1% resistance to chloramphenicol but only 1-5% (1/66) resistance to ampicillin. Resistance was 41-7% in our series. Have their laboratory methods failed to detect penicillin intermediate-resistant strains responsible for considerable morbidity and mortality? Their diffusion susceptibilty method (Pasteur Diagnostics) uses a 10 Irg ampicillin disc. A zone size of 29 mm is used to identify
438
intermediate and 21 mm to identify fully resistant strains. Jacobs et al showed in 19803 that these discs lack sensitivity and specificity, analysis of their histogram suggesting that 897%-966% of intermediately resistant pneumococci would be missed with a 10 µg ampicillin disc at a cut-off of 21 mm and at least 37-9% at 29 mm. If Pecoul et al used the former cut-off the great majority of intermediately resistant strains might not have been identified as such. A 1 µg oxacillin disk is the preferred method of
Because of the previous rarity of PRP, it has been usual practice in the UK to treat patients with proven pneumococcal meningitis with high doses of benzylpenicillin intravenously without waiting for antibiotic sensitivity results. It would now be preferable to initiate treatment with a third-generation cephalosporin such as cefotaxime. Therapy can be changed to penicillin if the strain is subsequently shown to be sensitive by appropriate testing.
evaluating pneumococcal susceptibiliry4 to penicillin/ampicillin. Chloramphenicol-resistant pneumococci were first isolated in west
Department of Medical Microbiology, Guy’s Hospital, London SE1 9RT, UK
Africa in 1978 but since that time few data, obtained with proper methods, are available on the prevalence of penicillin or ampicillinresistant strains in sub-Saharan Africa, north of South Africa. The data of Pecoul et al argue the need for information on the prevalence of penicillin-resistant pneumococcal strains in third world countries. Unfortunately, the third-generation cephalosporins required to treat these infections are unavailable in these countries on the grounds of cost. Department of Medical Microbiology, MRC Emergent Pathogen Research Unit, South African Institute for Medical Research, and University of the Witwatersrand, PO Box 1038, Johannesburg 2000, South Africa
KEITH P. KLUGMAN HENDRICK J. KOORNHOF
Department of Paediatrics, Baragwanath Hospital, Soweto
IAN R. FRIEDLAND
F, Keita M, et al. Long-acting chloramphenicol versus intravenous ampicillin for treatment of bacterial mengingitis. Lancet 1991, 338: 862-65. 2. Friedland IR, Klugman KP. Failure of chloramphenicol therapy in penicillin-resistant pneumococcal meningitis. Lancet 1992; 339: 405-08. 3. Jacobs MR, Gasper MN, Robins-Browne RM, Koornhof HJ. Antimicrobial susceptibility testing of pneumococci II. determination of optimal disc diffusion test for detection of penicillin resistance. J Antimicrob Chemother 1980; 6: 53-64. 4. Klugman KP. Pneumococcal resistance to antibiotics. Clin Microbiol Rev 1990; 3: 1. Pécoul B, Varaine
171-96.
SIR,- The prevalence of penicillin-resistant Streptococcus pneumoniae (PRP) varies greatly, resistance rates of 0% in France,’ 24% in South Africa,2 and 38% in Spain3having been reported. In the UK, PRP is uncommon but, as Ridgeway and colleagues have pointed out’ the rate is increasing and there was an almost four-fold rise in the number of PRP referred to the Central Public Health Laboratory from 1987 to 1991.5 We report here another case of PRP meningitis in which the organism was initially reported as penicillin sensitive. A 72-year-old woman was admitted to a London hospital in January, 1992, with symptoms of meningitis and a fever of 38.5°C. Patchy consolidation of the right lung field was seen on chest X-ray. The white blood cell count was 35 x 109/1 (92% neutrophils). A lumbar puncture yielded turbid cerebrospinal fluid (CSF) (white cell count 2-4 x 109/1 [95% neutrophils], protein 3-8 g/1, and glucose 01 rnmol/1 [peripheral blood glucose 7-4 mmol/1]). Gram-positive diplococci were seen on a CSF smear. The patient was admitted to intensive care, sedated, and intubated. She was put on intravenous ceftazidime 2 g every 6 h. S pneumoniae was grown from the CSF and was reported as sensitive to penicillin (1 µg penicillin disc), ampicillin, and ceftazidime. The organism was also grown from blood cultures. On day 2, the patient’s renal function appeared to be deteriorating and she was transferred to Guy’s Hospital. Antibiotic therapy was changed to benzylpenicillin 2 megaunits every 4 h. Pneumococci were again grown from blood cultures but on sensitivity testing, there was no zone around a 1 µg oxacillin disc. A computed tomographic scan of the brain revealed no abnormality. The patient recovered slowly and was discharged from intensive care 2 weeks later. The organism was a type 9 pneumococcus which had penicillin minimum inhibitory and bactericidal concentrations of 1 mg/1. This case illustrates the importance of testing isolates of S pneumoniae against a 1 Ilg oxacillin disc if penicillin resistance is to be detected. A study in 1987 showed that laboratories using oxacillin discs made fewer errors in detecting resistance or moderate resistance of pneumococci to penicillin than those using high content penicillin discs alone.6 This is especially important in meningitis, for which penicillin is inadequate as treatment even with moderately resistant strains (MIC 0’1-1 mg/1).
M. I. ISSACK N. A. SIMMONS G. L. FRENCH
AY, Goldstein FW, Acar JF. A seventeen year epidemiological survey of antimicrobial resistance in pneumococci in two hospitals. J Antimicrob Chemother 1988; 22 (suppl B). 41-52. 2 Klugman KP, Koornhof HJ, Gasas A, Storey K, Gilbertson I Carnage of penicillin-resistant pneumococci. Arch Dis Child 1986; 61: 377-81. 3. Casal J, Fenoll A, Vicioso M, Munoz R. Increase in resistance to penicillin in pneumococci in Spain. Lancet 1989; i: 735. 4. Ridgway EJ, Allen KD, Neal TJ, Lombard M, Rigby A Penicillin-resistant pneumococcal meningitis. Lancet 1992; 339: 931. 5. George RC, Ball LC, Cooper PG. Antibiotic resistant pneumococci in the United Kingdom. CDR Rev 1992; 2: 37-13. 6. Snell JJS, George RC, Perry SF, Erdman YJ. Antimicrobial susceptibility testing of Streptococcus pneumomae quality assessment results. J Clin Pathol 1988, 41: 1. Buu-Hoi
384-87.
SIR,-Recent communications on chloramphenicol in bactenal meningitis (predominantly pneumococcal) may cause anxiety among clinicians in the developing world, where the therapeutic options are already limited. We are often fortunate if we can identify the causative organism causing infection, let alone its antibiotic sensitivity, and the choice of antibiotic is often a matter of expediency rather than one based on hard micobiological data. Friedland and Klugman’s prospective studyl and the casereports of George et aP and Dr Corcoran and colleagues (July 4, p 50) describe initial treatments of up to 48 h with penicillin in combination with chloramphenicol and resultant poor responses in patients with penicillin-resistant pneumococcal meningitis. Penicillin was used in two instances even when resistance had been noted. It would have been helpful (especially in the prospective study) to know, in the presence of relative penicillin insensitivity, whether there was antagonism as measured by fractional lethal concentration indexes. In a wide range of organisms antagonism between penicillins and chloramphenicol has been described in which chloramphenicol itself was acting only as a bacteriostatic agent.3,4 The 48 % of children in the penicillin-resistant group who required alteration of antibiotic from chloramphenicol’ might represent a subgroup of patients who were relatively untreated during the first 48 h. Where bacterial meningitis has to be treated "blind" few clinicians with access to third-generation cephalosporins would not use them. Chloramphenicol alone in bacterial meningitis has a good record in developing countries5,6 and if penicillin resistance in
pneumococci were automatically to preclude chloramphenicol use, many adults and children would be in effect untreatable in many countries. The interaction of penicillin with chloramphenicol in
penicillin-insensitive organisms and the role of chloramphenicol as primary monotherapy need to be more clearly defined before there is a move away from chloramphenicol in the treatment of bacterial meningitis in developing countries. Department of Paediatrics, Royal Victoria Hospital, Banjul, The Gambia
THOMAS MCKAY
1. Friedland IR, Klugman KP Failure of 2.
chloramphenicol therapy in penicillin-resistant pneumococcal meningitis Lancet 1992; 339: 405-08 George RE, Wariyar UK, Nao GG. Penicillin-resistant pneumococcal meningitis
Lancet 1992; 339: 931-32. 3 Mathies AW, Leedom JM, Ivler D, et al. Antibiotic antagonism in bacterial meningitis. Anitmicrob Agents Chemother 1967, 7: 218-24. 4. Weeks JL, Mason EO, Baker CJ Antagonism of ampicillin and chloramphenicol for meningeal isolates of group B streptococci Antimicrob Agents Chemother 1981, 20: 281-85. 5. Shann F, Barker J, Poore P. Chloramphenicol alone versus chloramphenicol plus penicillin for bacterial meningitis m children Lancet 1985, ii: 681-84 6 McKay T Experience of changing antibiotic protocol in childhood bacterial meningitis in Vanuatu. Trop Doc 1989; 19: 158-59.
