Trimethoprim Resistance

1285

compared with 0-008 in a conventionally treated group.’ This rate in patients is similar to rates reported in other studies in comparison was made with conventional treatmentnamely 0’15,8 024,9 and 0-1210 episodes per patient per year. In subsequent years the rate of ketoacidosis in our CSII-treated patients fell, once the potential problem had been recognised and patients had been warned about ketoacidosis and re-educated in self-management at times of deteriorating blood glucose control.6 This reduction of ketoacidosis with increased experience of CSII is similar to that reported by Bending et al:l during the first years of experience of CSII the rate was 6 episodes per patient per year, falling to zero in 1984. Any "reassurance" drawn from studies showing that ketoacidosis rates in patients on CSII and in those on conventional treatment do the CSII which no

differ should be viewed with caution because these studies were done at centres with considerable experience of CSII and the lower rates of ketoacidosis emerged as physicians and patients gained experience of the treatment. In the report by Teutsch et all’ of deaths during CSII treatment 4 of 7 deaths attributed to ketoacidosis were in patients with less than 6 months’ experience of CSII. Physicians contemplating the use of CSII need to realise that the increased risk of ketoacidosis is greater during CSII than with conventional injection treatment, and they should pay special attention to patient selection and patient education.

difference between the groups for rates of ketoacidosis at 12 months. The high rate of ketoacidosis reported by the Kroc study was confined to two centres. Plasma hypertonicity influences transcellular potassium fluxes in both diabetics and non-diabetics.5 Plasma hypertonicity causes cellular dehydration, increased intracellular potassium concentrations, and increased outward passive potassium diffusion but there is no evidence that intracellular protein breakdown, insulin deficiency, or metabolic acidosis are important prerequisites.ED. L. no

1. Launtzen

2.

not

G. KNIGHT J. D. WARD

Royal Hallamshire Hospital, Sheffield S10 2JF

1. Bending JJ, Pickup JC, Keen H. Frequency of diabetic ketoacidosis and hypoglycemic coma during treatment with continuous subcutaneous insulin infusion. Am J Med 1985; 79: 685-91. 2. Sonnenberg GE, Muhlhauser I, Chantelau E, Berger M. Incidence of ketoacidosis and severe hypoglycemia in conventionally and CSII treated type I diabetic patients. Diabetes 1985; 34 (suppl 1): 117A. 3. Kroc Collaborative Study Group. Blood glucose control and the evolution of diabetic 1984; 311: 365-72. retinopathy and albuminuria. N Engl Med J 4. Pickup JC, Sherwin RS, Tamborlane WV, Rizza RA, Service F for the Kroc Collaborative Study Group. The pump life: Patient responses and clinical and technological problems. Diabetes 1985; 34 (suppl 3): 37-40. 5. Kashiwagi S, Wiefels K, Schleppinghof B, Gries FA, and DFI Group. Risk-benefit analysis of strict metabolic control in diabetics—the Dusseldorf prospective study. Diabetes Res Clin Pract 1985; 1 (suppl 1)· 5289. 6. Knight G, Boulton AJM, Drury J, et al. A feasibility study of the use of continuous subcutaneous insulin infusion m a diabetic clinic: patients’ choice of treatment. Diabetic Med 1984; 1: 267-72.

Knight G, Jennings AM, Boulton AJM, Tomlinson S, Ward JD. Severe hyperkalaemia and ketoacidosis during routine treatment with an insulin pump. Br Med J 1985; 291: 371-72. 8. Mecklenburg RS, Benson EA, Benson JW, et al. Acute complications associated with insulin infusion pump therapy. JAMA 1984; 252: 3265-69. 9. Peden NR, Braaten JT, McKendry JBR. Diabetic ketoacidosis during long term 7.

treatment

with continuous subcutaneous insulin infusion. Diabetes Care 1984; 7:

1-5. 10.

Irsigler K, Kritz H, Najemnik C, Fleissner M, Hagmueller G. Long-term treatment with external and implanted insulin infusion devices. Diabetes 1984; 33 (suppl 1):

73A. 11. Teutsch

SM, Herman WH, Dwyer DM, Lane JM. Mortality among diabetic patients using continuous subcutaneous insulin-infusion pumps. N Engl J Med 1984; 310: 361-68.

SIR,-Your editorial (Oct 11,

p

854)

seems

to

contain

a

fundamental misconception-namely, the view that potassium can be abstracted from cells by osmotic forces. That can only happen if the cell

proteins

are

being catabolised, which,

of course, is the

situation in insulin deficiency. Most of the potassium and sodium ions within a cell are bound to proteins, or are otherwise sequestered in the cytosol.1 E. N. WARDLE

Ling GN, Ochsenfeld MM. Sodium and potassium levels in living cells: do they depend on the outward transport of potassium? Physiol Chem Phys 1976; 8: 389-96.

* In

the

prospective study cited by Dr Knight

and Dr

Ward,

patients were not randomly allocated to CSIII or CIT. At least three other

investigationsi-3 with

4.

5.

DIABETES WITH KIDNEY FAILURE

SIR,-Professor Cameron and Dr Challah (Oct 25, p 962) review experience in treating uraemic diabetics in the UK in 1975-84 and note that in 1983-84 76% of newly accepted patients were type I. Continuous ambulatory peritoneal dialysis was the most popular long-term treatment (48 % of 446 under treatment at Dec 31,1984). The number of type I patients will be overestimated because the UK registry during the-years surveyed did not distinguish between a true insulin dependent diabetic (type I) and a type II patient treated with insulin by physician choice. We have studied this in our patients under treatment in fourteen dialysis centres, where 87 of 139 (62-6%) patients who are clearly type II diabetics have been or with insulin.1 From across the Atlantic, it seems that grave problems still exist in the British approach to uraemia therapy for the diabetic. Most diabetics with kidney failure in the United States are older type IIs-in Brooklyn, where type I patients constitute only 16-5% of diabetics undergoing maintenance haemodialysis, the mean age of 232 diabetics on dialysis on March 10, 1986, was 59-6years. The type II major patient subset appears to be excluded in the UK, a conclusion inferred from the lower mean age of 45years in treated patients reported by Cameron and Challah. In terms of survival and rehabilitation peritoneal dialysis and haemodialysis-the prevalent therapies in the UK- are inferior to transplantation.2 Although a rising proportion of uraemic patients accepted for treatment in the UK is diabetic (11’% in 1984, contrasting with 26 % in the USA) it is a sad reality that in the UK more than 50% of diabetics with kidney failure die without treatment. are on treatment

Renal Diseases Division, Department of Medicine, SUNY Health Science Center, Brooklyn, NY 11203, USA 1. Lowder

GM, Pern NA, Thiruvengadam tation and diabetes type in diabetics

ELI A. FRIEDMAN T,

Friedman EA

on maintenance

Employment, rehabilihemodialysis American

Society of Nephrology (abstr) (in press). EA, Peterson CM, eds. Diabetic nephropathy. Boston: Nijhoff,

2 Friedman

1986

TRIMETHOPRIM RESISTANCE

King Saud College of Medicine, Abha, Saudi Arabia 1.

3.

T, Frost-Larson K, Larson H-W, et al. Effect of 1 year of near-normal blood glucose levels on retinopathy in insulin-dependent diabetics. Lancet 1983; i: 200-04. Dahl-Jørgensen K, Brinchmann-Hansen O, Hansen KF, et al. Effect of nearnormoglycaemia for two years on progression of early diabetic retinopathy, nephropathy and neuropathy: the Oslo study Br Med J 1986; 293: 1195-99 Coustan DR, Reece A, Sherwin RS, et al. A randomised trial of the insulin pump vs intensive conventional therapy in diabetic pregnancies. JAMA 1986; 255: 631-36 DCCT Research Group. Results of feasibility phase of the Diabetes Control and Complications Trial (DCCT): glycemic control, follow-up and complications of therapy. Diabetes 1986; 35 (suppl 1): 3A. Moreno M, Murphy C, Goldsmith C. Increase in serum potassium resulting from the administration of hypertonic mannitol and other solutions. J Lab Clin Med 1969; 73: 291-98.

random

allocation

revealed

no

difference or a lower frequency of ketoacidisis on CSII versus CIT. In the feasibility phase of the Diabetes Control and Complications

Trial,4 in which 278 diabetics were randomly assigned to standard therapy or an intensified regimen which included CSII, there was

SIR,-Your Oct 4 editorial states that "the overall level of trimethoprim usage will have the greatest influence in the future on the incidence of trimethoprim resistance". The use of antibiotics does indeed cause resistance! but you do not answer the questionHow widely should trimethoprim be used? Trimethoprim resistance has been spread whether trimethoprim is used alone or in combination.2-4 If use of trimethoprim, alone or in combination, is concentrated in hospitals, especially geriatric hospitals, the spread of resistance is more than probable.4 We have seen this clearly in a geriatric hospital with annual consumption of trimethoprim of 5-10 g per bed. However, in outpatients the

1286

spread of trimethoprim resistance is not so clearly associated with consumption.5 In the Turku area of Finland, where about 05 g trimethoprim is used per inhabitant, trimethoprim resistance among outpatient strains of urinary tract Escherichia coli has been at a plateau of 10-12% for more than five years. In the Helsinki area, the consumption of trimethoprim is only about 02 g per inhabitant annually, but trimethoprim resistance among urinary tract E coli increased from 29% in 1980 to 11% in 1984 and to 14-5% in 1986.’*’* An explanation for this phenomenon is that the more crowded Helsinki area harbours more hospitals and geriatric nursing homes, which may have a role as a reservoir of trimethoprim resistance transposons. The "typical resistance level for Europe and North America" of 4-15% mentioned in your editorial is confusing without further explanation. These figures are, we assume, for urinary tract E coli. In Finland, trimethoprim resistance of E coli varies from 2-7% among strains collected from urine of outpatients less than 65 years old in the Rovaniemi area to 39% among strains collected from all patients in a geriatric hospital in the Turku area.4.5These figures reveal how important it is to characterise the patients involved, especially if the figures given may be used by clinicians in their choice of antimicrobial agents. Although trimethoprim is an important antimicrobial agent, we should not exaggerate the importance of resistance. Trimethoprim is seldom used in life-threatening infections, and resistant strains are rarely catastrophic. However, development of trimethoprim resistance has to be monitored carefully, and consumption of all antimicrobials must be restricted by education of physicians and by rules established for each hospital or region. Infectious Disease Unit, Department of Medicine, Massachusetts General Hospital, Boston, Massachusetts 02114, USA of Medical

Department University, Turku, Finland

PENTTI HUOVINEN

FREQUENCY OF S2 ALLELE

*Normolipidaemic-plasma cholesterol < 6 nnnol/fasting tnglycendes 7-6 mmol/1, LDL cholesterol > 5 mmot/1, tnglycendes < 2 0 mmol/l; hypertriglyceridaemic-LDL cholesterol < 4 7 mmol/l, tnglycendes > 3 3 rmnol/1. tDtfference from normoliptdaemic controls sigmficant (p < 0 005, cht squared test).

hypertriglyceridaemia and,

Microbiology,

Hughes VM, Datta N, Conjugative plasmids Nature 1984, 302: 725-26.

as

controls, 32 healthy, normolipid-

aemic volunteers with no family history of premature coronary heart disease. The 58 hypercholesterolaemic individuals were divided into two groups according to the presence or absence of tendon xanthomas since we regard the latter subgroup as having a definite diagnosis of familial hypercholesterolaemia. Our previous fmdings2 of a statistically significant association between the S2 allele and hypertriglyceridaemia in an English population were confirmed (table). More importantly, there was a significant association between this genetic marker and primary hypercholesterolaemia in patients without tendon xanthomas. This association may be of greater clinical relevance and may help to explain the increased incidence3of the apo-AI/CIII S2 allele in patients with coronary heart disease. It is not known how the as yet uncharacterised genetic defect linked to this allele predisposes to the different types of hyperlipidaemia. This work was supported by the British Heart Foundation.

Turku

1.

IN PATIENTS WITH HYPERLIPIDAEMIA

PAAVO TOIVANEN in bacteria of the

’pre-antibiotic’

era.

2. Editorial. Co-trimoxazole resistance. Lancet 1986; i: 364. 3. Murray BE, Alvarado T, Kim K-H, et al. Increasing resistance of trimethoprimsulpha-methoxazole among isolates of Escherichia coli m developing countries. J Infect Dis 1985; 152: 1107-13. 4. Huovinen P, Pulkkinen L, Helin H-L, et al. Emergence of trimethoprim resistance in relation to drug consumption in a Finnish hospital from 1971 through 1984. Antimicrob Agents Chemother 1986; 29: 73-76. 5. Huovinen P, Renkonen O-V, Pulkkinen L, et al. Trimethoprim resistance of Escherichia coli in outpatients in Finland after ten years’ use of plain trimethoprim. J Antimicrob Chemother 1985, 16: 435-41. 6. Huovinen P, Haikala O, Herva E, et al. Antimicrobial resistance of E. coli and Staphylococcus aureus Suom Laakaril 1986; 41: 2283-86 (in Finnish).

GENETIC MARKER IN APOLIPOPROTEIN AI/CIII GENE COMPLEX ASSOCIATED WITH HYPERCHOLESTEROLAEMIA

SIR,-In 1983 we reported an association between hypertriglyceridaemia and a restriction fragment length polymorphism associated with the apolipoprotein AI gene which was not present in normolipidaemic individuals. This polymorphism arises because of the existence of a polymorphic nucleotide in the 3’ non-coding region of the linked apo-CIIIgene 1

that creates an additional cleavage site for the restriction enzyme SacI.2 Two groups have since reported a significant increase in the incidence of the variant apo-AI/CIII allele (S2) in patients with severe coronary heart disease and in survivors of myocardial infarction compared with healthy controls.3,4 However, this increased frequency could not entirely be explained by a greater number of hypertriglyceridaemic individuals in these patients. In view of these observations and the fact that Kessling and Humphriess could not confirm our initial findings,’ we have tried to clarify the relationship between the S2 allele and hyperlipidaemia by studying newly recruited patients with well defmed forms of

hyperlipidaemia. We assessed the frequency of the S2 allele2 in 58 unrelated adults with primary hypercholesterolaemia, 26 patients with primary

Sir William Dunn School of Pathology, University of Oxford, Oxford OX1 3RE; and Lipid Clinic, John Radcliffe Hospital, Oxford

C. C. SHOULDERS M. J. BALL J. I. MANN F. E. BARALLE

1. Rees A, Shoulders CC, Stocks J, Galton DJ, Baralle FE. DNA polymorphism adjacent to human apoprotein A-1 gene: relation to hypertriglyceridaemia Lancet 1983; i. 444-46. 2. Shoulders CC, Baralle FE. Genetic polymorphisms in the apo-AI/CIII complex Methods Enzymol 1986; 128 (part A): 727-45. 3. Rees A, Stocks J, Williams LG, et al. DNA polymorphisms in the apolipoprotein CIII and insulin genes. Atherosclerosis 1985, 58: 269-75. 4. Ferns GAA, Stocks J, Ritchie C, Galton DJ. Genetic polymorphisms of apolipoprotein CIII and insulin in survivors of myocardial infarction Lancet 1985, ii: 300-03. 5. Kessling AM, Humphries SE. Interpretation of presence of S2 allele. Lancet 1985, ii 510.

SIR,-Dr Price and his colleagues (Nov 1, p 1041) have not found an uncommon allelic variant of the locus (the S2 allele) and coronary artery apo-AI-CIII-AIV gene disease (CAD) in a Scottish population. The restriction fragment length polymorphisms (RFLP) of the apo-AI-CIII-AIV gene cluster investigated by our group and others" are probably acting as linkage markers for CAD. Such markers would be expected to show racial and regional differences. Although the S2 allele may not relate to CAD in the Scottish population, other RFLPs of the locus may demonstrate significant disease association within this population. Indeed Price et al themselves have found an association between a PstI RFLP at this locus and CAD within their Scottish samplesthough they do not mention this in their letter. an

association between

Department of Lipid Research, St Bartholomew’s Hospital, London EC1A 7BE 1. Ordovas

JM, Schaeffer EJ, Salem D,

G. A. A. FERNS

J. STOCKS D. J. GALTON

et al Apolipoprotein AI gene polymorphism disease and familial coronary artery premature hypoalphalipoproteinemia. N Engl J Med 1986; 314: 671-67 2. Rees A, Caplin JL, Jowett NI, et al. DNA polymorphism in the apo CIII and insulin gene and atherosclerosis Atherosclerosis 1985, 58: 269-75. 3 Fems GAA, Galton DJ. Haplotypes of the human apoprotein AI-CIII-AIV gene cluster in coronary atherosclerosis Hum Genet 1986, 73: 245-49.

associated

with