4
5
Rogier C, Trape JF. Malaria attacks in children exposed to high transmission: who is protected? Trans R Soc Trop Med Hyg 1993; 87: 245-46. Trape JF, Peelman P, Morault-Peelman B. Criteria for diagnosing clinical malaria among a semi-immune population exposed to intense and perennial transmission. Trans R Soc Trop Med Hyg 1985; 79: 435-42.
most of the potentially severe We also agree that the vaccine-related reduction in the number of episodes recorded could be due to a decrease in the intensity of symptoms rather than number of the attacks-ie, shifting the spectrum of clinical severity towards the more benign presentations. If this is so, the vaccine might have a substantial impact on mortality that needs to be documented.
the
episodes detected include
cases.
Authors’reply *Pedro L Alonso, Tom Smith, Marcel Tanner
and Rogier rightly point out that in the SPf66 trial the reported annual incidence of Tanzanian clinical episodes was lower than the number of episodes expected in an area of such high malaria transmission. There are probably several reasons for this. The highest risk of clinical malaria in the Kilombero population occurs around the first birthday. In areas like this one, where transmission is so intense, the frequency of clinical malaria declines very rapidly with age. Indeed 1-year-old children have twice the risk of 2-year-olds. For regulatory reasons we could not vaccinate children less than 1 year of age. Consequently, most of the time, at risk refers to children who were older (up to 7 years by the end of the trial) and had much lower risk of symptomatic episodes. Furthermore, the study cohorts were ageing as the trial progressed and the risk of malaria therefore decreased. At the same time, primary data analysis depended on passive case detection at the village dispensary. Even though the dispensary staff and supplies were re-enforced during the study period, and free medical services were offered around the clock for the entire duration of the trial, it is quite likely that children with mild, short-duration, and self-limited episodes did not attend the dispensary for consultation, and were therefore not reported. The dispersed structure of the village could have also contributed to the low recorded frequency (mean distance from the dispensary 1-9 km; interquartile range 0-9-4-8). The recorded frequency was higher in children who lived nearest the dispensary, suggesting that reporting rates were lower from more remote parts of the village. Moreover, symptomless infections occur at a high frequency in the Tanzanian study area, and this complicates the analysis of any trial of interventions against malaria morbidity. Children with fevers of non-malarial aetiology can have concurrent symptomless parasitaemia, and therefore we can never be entirely sure that the parasites are the cause of the fever; this is especially so when sick children are not necessarily examined at either the peak of fever or peak parasitaemia of the episode. In the Kilombero vaccine trial, many such episodes would not, at the time of sampling, have satisfied our primary case definition, which included only episodes presenting with an axillary temperature of 37-5°C or more and a parasite density greater than 20 000/,L. The risk of fever does, however, increase with the density of parasitaemia. We have lately used this observation to show how the sensitivity and specificity of different parasitaemia cutoffs can be determined.’2 The case definition used for the primary analysis was established before the randomisation code was broken, with the data obtained during the 6 months of follow-up between the first and third dose. A definition with a high specificity was chosen to avoid the dilution of the vaccine effect by inclusion of fever cases with non-malarial aetiology. All these considerations combined with the known timedependent variability on the frequency of clinical malaria as indicated by the reduction over time in the age-standardised rates of malaria morbidity in both the vaccine and placebo cohorts, might have contributed to the lower-than-expected incidence of malaria recorded in the Kilombero vaccine trial. Finally, we agree with Trape and Rogier that it is likely that
SIR-Trape
*Unidad de Epidemiologia y Bioestadistica, Hospital Clinic i Provincial, 08036 Barcelona, Spain; and Department of Epidemiology and Public Health, Swiss Tropical Institute, Basel, Switzerland
1
2
Smith TA, Armstrong Schellenberg JRM, Hayes R. Attributable fraction estimates and case definitions for malaria in endemic areas. Stat Med 1994; 13: 2345-58. Armstrong Schellenberg JRM, Smith T, Alonso PL, et al. What is clinical malaria? Finding case definitions for malaria in endemic areas. Parasitol Today 1994; 10: 429-42.
Regional variations in medium-chain acyl-CoA dehydrogenase deficiency SiR-Medium-chain acyl-CoA dehydrogenase (MCAD) disorder of is an autosomal recessive mitochondrial fatty acid oxidation. About 90% of patients are homozygous for the same mutation (G985). Affected individuals are at risk of metabolic decompensation, especially during infancy, and may present with hypoglycaemia, encephalopathy, hepatomegaly, and sudden death or near-miss episodes, though some remain symptomfree.’ Preliminary studies from the West Midlands2 and Trent3 regions of the UK found heterozygosity for G985 in 12 of 479 and 6 of 410 babies, respectively, suggesting that in these areas 1 in 10000 individuals have MCAD deficiency and supporting anecdotal evidence that this is a relatively common disorder, and one that may be widely underdiagnosed.’ However, the statistical limitations of these small surveys left unanswered many questions about the impact of the disease and the case for routine neonatal screening. We have now examined over 10 000 samples and have compared the predicted frequencies of MCAD
deficiency
deficiency with the
rates of diagnosis in the two regions. The West Midlands study covered the Herefordshire, Shropshire, and North and South Worcestershire health districts. In Trent eleven of the thirteen districts participated. Dried blood surplus to routine screening was tested for G985 by polymerase chain amplification and allele-specific hybridisation.’ All samples in the heterozygous range were then examined by Styl restriction enzyme digestion followed by gel electrophoresis. (Details of assay performance will be reported elsewhere but if there is any bias it will be to underestimate heterozygote frequency.) The results (table) confirm previous indications2°3 that gene frequency is higher in the selected West Midlands districts than in Trent and that there is significant regional variation in the frequency of MCAD deficiency. A lower frequency of G985 MCAD mutation has been reported from the west of Scotland.’5 There has been a long-standing interest in MCAD deficiency in Sheffield (Trent) and a high index of suspicion is to be expected. The rate of diagnosis in the West Midlands is much lower, although using the gene frequency from a selected area to calculate the expected number of cases for the whole region may have introduced some bias. Explanations for the apparently missing cases in the West Midlands could include a lower awareness of MCAD deficiency and its presenting features, failure to detect the
135
drops. The toxicity of chlorhexidine for epithelial cells (skin, conjunctival, and corneal), fibroblasts, and other cells
eye
concentrations equal to or greater than 0-2% w/v is well but for mucosal infection 0-02% (200 llg/mL)4 is safe. This concentration is effective against acanthamoeba cysts; it is also bactericidal since it induces membrane disruption. At this concentration mammalian cells are unaffected and there is no apparent ocular toxicity. In vitro it is the most active of the biguanides,I,2 with an average minimum cysticidal concentration of 2 g/mL. PHMB (Zeneca Specialities) is not manufactured to GMP standards. It is used as an environmental disinfectant under trade names such as Cosmocil CQ, Vantocil IB, and Baquacil, varying in purity. Baquacil is available as a swimming pool disinfectant in the USA; it is an effective sanitiser and is algistatic and amoebicidal. Cosmocil CQ is used for the preservation of cosmetics and pharmaceuticals, and PHMB (polyhexanide) is in some soft-contact-lens disinfectant solutions. PHMB is not manufactured or licensed for therapeutic use. It is generally less effective than chlorhexidine in vitro against acanthamoeba cysts.1,2 PHMB for acanthamoeba keratitis is thus essentially an experimental treatment. We think that it should be replaced by chlorhexidine. It also appears more rational, to protect against the emergence of resistance, that PHMB be restricted to environmental disinfection. For some purposes, such as disinfection of urinary catheter drainage bags, chlorhexidine might be replaced by PHMB but the ophthalmological application of biguanide antiseptics should be restricted to chlorhexidine. Elder et al have described two treatment failures with PHMB in which the post-treatment keratoplasty specimen yielded Acanthamoeba spp that were fully sensitive to PHMB in vitro." This has not yet been recognised with chlorhexidine. For acanthamoeba keratitis we recommend 0-02% chlorhexidine digluconate drops in 0-9% physiological saline in combination with 0-1% propamidine isethionate drops, given hourly day and night for the first 3 days, then 2-hourly by day for 4 weeks, 3-hourly by day for 4 weeks, and 4hourly by day for 4 months.’ Compliance is essential and this can be a problem in contact lens wearers who had acquired their infection by non-compliance with lens disinfection.
at
recognised4
tcalculated on basis of 90% of disease-producing mutations being G985;‡ln genotyped, 94% of mutations were G985. Table: Calculated and observed frequencies of MCAD *p<001;
24 families
deficiency characteristic
urine
metabolites
during a symptomless period, symptom-free cases. The
in and
specimens collected a large number of
for routine neonatal screening for MCAD hinges the status of the missing cases: are they large or symptom-free undiagnosed? The Trent data suggest that even when ascertainment is almost complete most MCADdeficient children have symptoms: in all the 18 families an affected child had either died or experienced an acute lifethreatening episode. A prospective survey of the diagnosis of MCAD deficiency is being undertaken by the British Paediatric Surveillance Unit and this, taken with gene frequency data of the type presented here, will help to define the impact of the disorder nationally.
to a
case
extent on
Supported by the National Reyes’ Syndrome Foundation. We also thank Dr P Dahlen and Wallac Oy, Turku, Finland, for guidance with the assay and the gift of DELFIA reagents. H R Seddon, A Green, R G F
Gray, J
V
Leonard, *R J Pollitt
Department of Clinical Chemistry, West Midlands Regional Laboratory for Inherited Metabolic Disease, Children’s Hospital, Birmingham; Institute of Child Health, London; and *Neonatal Screening Laboratory, Children’s Hospital, Sheffield S10 2TH, UK 1
Editorial. Medium-chain acyl-CoA
2
1991; 338: 544-45. Matsubara Y, Narisawa K, Tada K,
3
4
5
dehydrogenase deficiency. Lancet
et al. Prevalence of K329E mutation in medium-chain acyl-CoA dehydrogenase gene determined from Guthrie cards. Lancet 1991; 338: 552-53. Blakemore AIF, Singleton H, Pollitt RJ, et al. Frequency of the G985 MCAD mutation in the general population. Lancet 1991; 337: 298-99. Iitiä A, Mikola M, Gregersen N, Hurskainen P, Lövgren T. Detection of a point mutation using short oligonucleotide probes in allele-specific hybridisation. Biotechniques 1994; 17: 566-73. Dundar M, Lanyon WG, Connor JM. Scottish frequency of the common G985 mutation in the medium-chain acyl-CoA dehydrogenase (MCAD) gene and the role of MCAD deficiency in sudden infant death syndrome (SIDS). J Inher Metab Dis 1993; 16: 991-93.
*D V Seal, J
1
2
Chlorhexidine or polyhexamethylene biguanide for acanthamoeba keratitis SiR-Two biguanide compounds-chlorhexidine digluconate and polyhexamethylene biguanide hydrochloride (PHMB)are recommended for treatment of acanthamoeba keratitis.’-3 Both are acanthamoebicidal’ and both are effective in patients with this corneal infection, when used in combination with Brolene (0-1% propamidine isethionate) as frequent drop therapy. Why then is there debate over which compound to use? Chlorhexidine (Zeneca Pharmaceuticals) is a licensed
good manufacturing practice (GMP) standards. It has been used successfully for 30 years to treat bacterial skin infection, open wounds, and gingivitis; it is also effective for antiseptic handwashing and surgical skin preparation. The diacetate has been used as a bactericide in antiseptic produced
136
Hay, C
M Kirkness
Tennent Institute of Ophthalmology, University of Glasgow G11 6NT, UK
3
4
5
Glasgow,
Western
Infirmary,
Hay J, Kirkness CM, Seal DV, Wright P. Drug resistance and Acanthamoeba keratitis: the quest for alternative antiprotozoal chemotherapy. Eye 1994; 8: 555-63. Elder MJ, Kilvington S, Dart JKG. A clinicopathologic study of in vivo sensitivity testing and Acanthamoeba keratitis. Invest Ophthalmol Vis Sci 1994; 35: 1059-64. Varga JH, Wolf TC, Jensen HG, Parmley VC, Rowsey JJ. Combined treatment of Acanthamoeba keratitis with propamidine, neomycin and polyhexamethylene biguanide. Am J Ophthalmol 1993; 115: 466-70. Archer HG, Barnett S, Irving S, Middleton KR, Seal DV. A controlled model of moist wound healing: comparison between semi-permeable film, antiseptics and sugar paste. J Exp Pathol 1990; 71: 155-70. Seal DV, Kirkness CM, Hay J, Morell A. Successful medical treatment of Acanthamoeba keratitis with chlorhexidine eye drops. 28th annual meeting of Ocular Microbiology and Immunology Group (San Francisco, Oct 29, 1994); abstr 53.
CORRECTION
to
Learnmg about HIV-2-In this Commentary by M V O’Shaughnessy and M T Schechter (Nov 19, p 1380) reference 8 was inadvertently omitted from the list. This should read Marlink R, Kanki P, Thior I, et al. Reduced rate of disease development after HIV-2 infection as compared to HIV-1. Science 1994; 265: 1587-90.