279

patients in both studies was different; the average in John-stone’s study was 57 (range 42-70), and in ours was 34 (range17-66), only three of our patients being over 40. Since thereis an increased incidence of cerebral atrophy with age it is dif-ficult to assess how great a part this factor has played in theresults of Johnstone’s study. Thirdly, the diagnosis of schizo-phrenia is always open to doubt and it is possible that some pa-tients in both studies were incorrectly assessed.To confirm the findings of Johnstone et al. studies must be

done with patients in whom a diagnosis of schizophrenia canbe substantiated by a technique such as present state examina-tion. Our normal findings in a younger group of schizophrenicpatients mean that several factors need consideration, namelythe development of progressive cerebral atrophy from neuro-leptic drugs,3 chronic institutionalisation,4 and dietary fac-tors.4 There is also the possibility that the patients studied byJohnstone et al. were a subgroup, possibly having some accom-panying dementing process.

It is clear that not all patients with schizophrenia haveabnormal ventricular size but our results, taken in conjunctionwith those of Johnstone et al., raise the possibility that cerebralatrophy develops as the disease progresses. If this is so, the rea-sons for it deserve investigation.Departments of Psychological Medicine,and Diagnostic Radiology,

National Hospital,London WC1N 3BG

MICHAEL TRIMBLEDEREK KINGSLEY

SEX-SPECIFIC CARCINOGENS, WITH SPECIALREFERENCE TO SACCHARIN

SIR,-One of your criticisms’ of the report by Howe et al.’ 6of a positive association between the use of saccharin and blad-der cancer among men but not women was that no known non-hormonal carcinogen affecting man is sex-specific. Theabsence of such evidence is not a strong argument; further-more, sex-specific non-hormonal carcinogens are well knownfrom studies of carcinogenesis in animals. A review of volumes-xv of the IA.R.C. Monograph series revealed 22 carcinogenswhich are sex-specific (incidence of tumours in one sex was atleast 3 times that in the opposite sex) in at least one strain ofanimals; males were more susceptible than females in 19 of the22 instances. The number of sex-specific carcinogens bytumours produced were: hepatomas (11), lymphomas (3), largeintestine (2), bladder (1), pancreas (1), vascular (1), lung (1),sarcoma (1), and mixed (1). A similar pattern in man wouldhardly be surprising.Cancer Section,Bureau of Epidemiology,Laboratory Centre for Disease Control,Health and Welfare Canada,Ottawa, Ontario K1A 0L2, Canada DONALD T. WIGLE

CIMETIDINE AND LARGE-BOWEL FUNCTION

SIR,-Watson et aU link cimetidine with ileus in the burnedpatient and mention the possibility that H2 inhibitors blockfunctions other than gastric acid secretion.My interest in this finding arose when I had an irritable

colon. Investigations included an intravenous pyelogram, anair contrast barium meal, three cholecystograms includingtomography, a gastroscopy, and a barium meal follow

through, none of which gave any evidence of organic disease.The symptoms were considered to fit with an irritable colon.

3. Marsden, C. D. Lancet, 1976, ii, 1079.4. Jellinek, E. H. ibid. p. 1202.5. Lancet, 1977, ii, 592.6. Howe, G. R., and others ibid. p. 578.7. Watson, W. C., Kutty, P. K., Colcleugh, R. G. Lancet, 1977, ii, 720.

Before this diagnosis duodenal ulcer was suspected and I hadbeen taking cimetidine (300 mg 4 times daily). This gave meconsiderable relief, a result which supports the idea that cime-tidine affects large-bowel function.

I have since given cimetidine to two or three patients inwhom intensive investigation revealed no significant lesion andin whom the most likely diagnosis was an irritable colon; allthe patients noted a subjective improvement while less im-provement followed other agents, particularly anti-spasmodics.Perhaps the use of cimetidine in the treatment of irritablecolon merits a properly controlled trial.

Apartment 503, Hunt Building,Lethbridge, Alberta, Canada EUAN M. S. FREW

SERUM-FERRITIN IN DIAGNOSIS OFHÆMOCHROMATOSIS

SIR,-Leyland et al.l found widely varying serum-ferritinconcentrations in patients with idiopathic haemochromatosisduring venesection therapy. The results contradict earlierobservations of Prieto et al. and our own experience. In six pa-tients with idiopathic haemochromatosis, serum-ferritin de-clined during venesection without major fluctuations (seefigure). Apart from one very high value late in the course oftreatment in patient A (see inset in figure), the variationsobserved were within the error of the method.3 A progressivedecline in serum-ferritin with little week-to-week fluctuation

1. Leyland, M. J., Brown, P. J., Walker, R. J., Bomford, A., Williams, R. Lan-cet, 1977, ii, 1030.

2. Prieto, J., Barry, M., Sherlock, S. Gastroenterology, 1975, 68, 525.3. Luxton, A. W. and others. Clin. Chem. 1977, 23, 683.

Serial serum-ferritin values in patients undergoing venesectiontherapy.The size of body iron stores was determined from the amount of

hamoglobin iron removed, corrected for an estimated daily absorptionof 5 mg of dietary iron during the treatment period.5

280

was also observed in one patient with cirrhosis of the liver andexcess stainable hepatocellular iron and in three relatives ofpatients with idiopathic haemochromatosis who were givenvenesection therapy because more than 50% of hepatocytes insections of a percutaneous liver biopsy specimen containedstainable iron (see figure). Three of our patients (A, C, D)drank generous amounts of beer before treatment but did notdrink alcoholic beverages to excess during phlebotomy.

In the three relatives with normal serum-ferritin concentra-tions, 1 ng/ml of serum-ferritin corresponded to approximately20 mg (16-26) of storage iron calculated from either the initialferritin value or from the rate of decline in the serum-ferritinconcentration. The relationship between the initial ferritinconcentration and the estimated size of iron stores was less pre-cise in the patients with a moderate to marked increase in ser-um-ferritin. Three patients with ferritin values of 1200 ng/ml,2116 ng/ml, and 1250 ng/ml had estimated body iron storesof 5 g, 5 g, and 16 g respectively. The rate of decline of serum-ferritin concentration was about the same in five of the pa-tients with idiopathic hxmochromatosis: a fall of 1 ng/ml ofserum-ferritin corresponded to approximately 2-5 mg(1-3-4-6) of storage iron. The rate of fall of serum-ferritin inthe remaining patient (A) with idiopathic hsemochromatosiswas slow until iron stores were nearly exhausted.Leyland and coworkers do not refer to the method used for

their analysis of serum-ferritin nor to its precision for knownsera assayed concurrently with patients’ sera. They also fail toindicate whether sera from individual patients were assayedwith the same set of standards. It is also important to knowwhether sera from patients were analysed regularly at morethan one dilution. Green et al. have alluded to the potentialpitfall of the high-dose hook effect in the serum-ferritin assay-sera from patients with severe iron-storage disease mis-takenly appearing to have low serum concentrations by a two-site immunoradiometric assay unless they are assayed at highdilutions.

Further information on the precision of the technique is

required before the results reported by Leyland et al. can beconfidently ascribed to biological variations in serum-ferritinconcentration during venesection.

Department of Medicine,University Hospital,

and University of Western Ontario,London, Ontario, Canada N6A 5C1

L. S. VALBERGD. MACKINNONJ. W. D. MCDONALD

HYPERGLYCÆMIC EFFECT OF SYNTHETICSALMON CALCITONIN

SIR,-Dr Gattereau and his colleagues’ have reported thata single subcutaneous injection of 100 M.R.C. units of syn-thetic salmon calcitonin resulted in hypergtycsemia, and theysuggest that long-term administration of calcitonin may lead todiabetes.We have treated over forty patients with Paget’s disease of

bone for up to 8 years with daily injections of synthetic humancalcitonin and none has clinical diabetes. Ziegler et al. 2

reported that the impairment of glucose metabolism seen afteran acute injection of calcitonin was not borne out in the longerterm. It is very unlikely that calcitonin treatment will causediabetes.

Endocrine Unit,Royal Postgraduate Medical Schooland Hammersmith Hospital,

London W12 0HS

IMOGEN M. A. EVANS

G. F. JOPLINIAIN MACINTYRE

4. Green, R. and others. Blood, 1977, 50, 545.5. Crosby, W. H., Conrad, M. E., Wheby, M. S. ibid. 1963, 22, 429.1. Gattereau, A., Bielmann, P., Durivage, J., Larochelle, P. Lancet, 1977, ii,

1076.2. Ziegler, R., Holz, G., Raue, F., Minne, H., Delling, G. in Human Calcitonin

and Paget’s Disease (edited by I. MacIntyre); p. 167. 1977.

DEMONSTRATION OF DAPSONE IN URINE ANDSERUM BY ELISA INHIBITION

S!R,—Conventional methods for demonstrating dapsone areeither insensitive or not practicable for general use in leprosyendemic areas. We have developed an immunoassay that isboth simple and sensitive. ,

Diazotised dapsone was conjugated to bovine serum albu-min, and 10 mg of conjugate emulsified in Freund’s completeadjuvant was injected into rabbits. Booster injections of 10 mgof conjugate in Freund’s incomplete adjuvant were given after3 weeks. Anti-dapsone antibody was demonstrated by micro-scale enzyme-linked immunosorbent assay (micro-ELISA). Weused diazotised dapsone/horseshoe-crab haemocyanin conju-gate (D.D.S-H.C.H.) to coat the wells of polystyrene microtitretrays, and horseradish-peroxidase-conjugated anti-rabbit IgGantiserum (Institut Pasteur, Paris). Sera positive for anti-dap-sone antibody were identified by adding a solution of 5-amino-salicylic acid/H202 and looking for. development of a browncolour. Addition of as little as 0.1 fLg dapsone to each well inhi-bited positive reactions by two doubling dilutions. 10 pg sul-phadiazine, a structural analogue of dapsone, did not inhibitthe ELISA response to the same extent. A test for inhibition ofELISA (ELISIT) was then developed to detect dapsone in urineand sera from leprosy patients. Urine was collected from 44leprosy patients in Kenya, 40 of whom were supposed to betaking 300 mg of the drug once a week, whilst the other 4 weretaking lower doses. -

The urine was collected 2 days after a scheduled dose. It waspreserved by adding hydrochloric acid. Before testing the urineits pH was adjusted to 7.2. In addition urine taken 6 h afterthe intake of 300 mg dapsone was obtained from one Dutchvolunteer. Urine from 16 healthy individuals not taking dap-sone were also tested. Sera of 3 leprosy patients taking 100 mgdaily and one volunteer whose blood-sample was taken 6 hafter a single dose of 300 mg dapsone were examined too. Con-trol sera from 4 healthy people not taking dapsone weretreated in the same way.The wells of the tray were first coated (30 min, 37 °C) with

100 1 of carbonate-buffered solution (pH 9-6) containing 0.2fLg D.D.S.-H.C.H. After the normal washing cycle, 50 ,1 of thesamples under test were added. The samples were diluted inP.B.s./tween/H.c.H. and 50 }jd of 1/500 diluted anti-dapsonerabbit serum was added. After incubation, washing, conjugate

Demonstration of dapsone by ELISIT.(1) and (2) A-H standard dapsone solutions of 10, 3, 1, 0.3, 0-1, 0.03,

0-01, and 0.003 fig/ml.(3)-(7) A-H and (8) A-D: urine from Kenyan leprosy patients, 1/10.(8) E-H: urine from Dutch volunteer, 1/10, 1/50, 1/250, and 1/1250.(9) and (10) A-H: urine from 16 healthy persons, 1/10.(11) A-F: sera from 3 Dutch leprosy patients (A and B, C and D, E

and F duplicates), 1/100.(11) G and H: serum from Dutch volunteer (duplicates), 1/100.(12) A-H: sera from 4 healthy persons (A and B, C and D, E and F,G and H duplicates), 1/100.