1178
not acquire antibodies transplacentally. Rotavirus anti-
bodies present in the sera of almost all human adults
may protect the newborn after transplacental transmis-
sion.Although it may appear surprising that serum anti-
body rather than secretory antibody protects against in-
fection localised in the small intestine, vaccine-induced
circulating antibodies protect the gut against cholera12
and the respiratory tract against influenza." However,
although newborn infants are protected from disease,
they are not protected from infection: 33% in our study
excreted virus, occasionally in very large amounts.
Symptomless infection with excretion of large quantities
of virus also occurs with cytomegalovirusi4 and hepati-
tis-B virus. Holmes has suggested that the intestinal
brush-border enzyme, lactase, acts as the host-cell recep-
tor for rotavirus.16 In-vitro trypsin treatment enhances
the in-vitro infectivity of porcinel7 and bovine18,19 rota-
viruses. However, there is no evidence that human
neonates are deficient in these enzymes.
It is possible that the rotavirus strains circulating in
our nurseries were avirulent. Pig rotaviruses may vary
in virulence.2o However, the demonstrations of mild or
symptomless infection in newborn infants but frank
diarrhoea in older children in the U.K., Australia,’O and
India2I do not support this hypothesis. Furthermore, in
our hospital there is concurrent rotavirus infection
among symptomless newborn infants and among older
children with gastroenteritis in the ward only one floor
above the newborn nurseries.
Breast-fed babies excreted rotavirus significantly less
frequently and, when infected, generally shed less virus
than bottle-fed babies. Breast-fed babies may be pro-
tected by maternal secretory antibodies in colostrum 22,23
and breast milk24 during the newborn period, or by non-
specific antiviral factors distinct from antibody or inter-
feron in breast milk. 25
In a rural Bangladesh environment heavily contami-
nated with rotaviruses admission to hospital with rotavir-
us-induced diarrhoeal disease is rare before the age of 6
months.26 How rotaviruses induce diarrhoea in humans
is not clearly understood but the newborn’s gastrointes-
tinal tract may have physiological characteristics not
found in older children which allow symptomless infec-
tion. Discovery of the mechanism by which young in-
fants are protected might provide an alternative to vac-
cination.
Although it is uncertain how infection was introduced
into the newborn nurseries, transmission of infection by
medical or nursing staff from infected older children in
the general wards seems likely. Studies in Toronto
have shown that rotavirus infection, a common cause of
hospital-acquired gastroenteritis among older children,
was probably transmitted by medical staff.27 We found
infection occasionally in infants in the newborn special-
care nursery. Premature babies also had symptomless in-
fections but in this unit many are in incubators and
active measures are taken to prevent cross-infection, so
rotavirus spread is likely to be rare.
We are now investigating whether our newborn rota-
virus-infected infants are "immunised" as a result of in-
fection. Follow-up studies on 20 babies who excreted
rotavirus neonatally have shown that 1 had rotavirus-
induced diarrhoea when aged 14 months, and we are
attempting to determine whether the two virus strains
are antigenically related. Serotyping by ELISA of
groups of 3 rotavirus-positive stool samples from the
1975 and 1976-77 surveys of neonatal nurseries and
from current episodes of diarrhoea among older children
has been carried out by Dr R. H. Yolken, N.I.H., Beth-
esda, U.S.A., and has shown that all viruses are type 2.
This type appears to be the predominant serotype associ-
ated with symptomatic disease in most parts of the
world.
We thank the medical officers and nursing staff of the maternity
wards at St. Thomas’s Hospital for their cooperation.
REFERENCES
1. Davidson, G. P., Bishop, R. F., Townley, R. R. W., Holmes, I. H., Ruck,
B. J. Lancet, 1975, i, 242.
2. ibid. p. 257.
3. Mebus, C. A., Underdahl, N. R., Rhodes, M. B., Twiehaus, M. J. Univ.
Nebraska Agric. exp. Station Res. Bull. 1969,no. 233.
4. Woode, G. N., Bridger, J. C. Vet. Rec. 1975, 96, 85.
5. Commun. Dis. Rep. publ. Hlth Lab. Serv. 1977, no. 30.
6. Chrystie, I. L., Totterdell, B., Baber, M. J., Scopes, J. W., Banatvala, J. E.
Lancet, 1975, ii, 79.
7. Totterdell, B. M., Chrystie, I. L., Banatvala, J. E. Archs Dis. Childh. 1976,
51, 924.
8. Bishop, R. F., Hewstone, A. S., Davidson, G. P., Townley, R. R. W.,
Holmes, I. H., Ruck, B. J., J. clin. Path. 1976, 29, 46.
9. Cameron, D. J. S., Bishop, R. F., Davidson, G. P., Townley, R. R. W.,
Holmes, I. H., Ruck, B. J. Med. J. Aust. 1976, i, 85.
10. Murphy, A. M., Albrey, M. B., Crewe, E. B. Lancet, 1977, ii, 1149.
11. Mebus, C. A., Stair, E. L., Underdahl, N. R., Twiehaus, M. J. Vet. Pathol.
1971, 8, 490.
12. Pierce, N. F. in Acute Diarrhoea in Childhood. (edited by K. Elliot and
J. Knight). Ciba Foundation Symposium no. 42; p. 129. Amsterdam,
1976.
13. Edmonson, W. P., Rothenburg, R., White, P. W., Gwaltney, J. M., Jr. Am.
J. Epidem. 1971, 93, 480.
14. Starr, J. G., Bart, R. D., Jr., Gold, E. New Engl. J. Med. 1970, 282, 1075.
15. Schweitzer, I. L., Spears, R. L. ibid. 283, 570.
16. Holmes, I. H., Ruck, B. J., Bishop, R. F., Davidson, G. P. J. Virol. 1975,
16, 937.
17. Theil, K. W., Bohl, E. H., Agnes, A. G. Am. J. vet. Res. 1977, 38, 1765.
18. Babiuk, L. A., Mohammed, K., Spence, L., Fauvel, M., Petro, R. J. clin.
Microbiol. 1977, 6, 610.
19. Almeida, J. D., Hall, T., Banatvala, J. E., Totterdell, B. M., Chrystie, I. L.
(In the press).
20. Woode, G. N. in Acute Diarrhoea in Childhood. (edited by K. Elliot and
J. Knight). Ciba Foundation Symposium no. 42; p. 251. Amsterdam,
1976.
21. Mathan, M. Unpublished.
22. Simhon, A., Mata, L. Lancet, 1978, i, 39.
23. Inglis, G. C., Sommerville, R. G., McClelland, D. B. L. ibid. p.559.
24. Thouless, M. E., Bryden, A. S., Flewett, T. H. Br. med. J. 1977, ii, 1390.
25. Matthews, T. H. J., Nair, C. D. G., Lawrence, M. K., Tyrrell, D. A. J.
Lancet, 1976, ii, 1387.
26. Yolken, R. H. Unpublished.
27. Middleton, P. J., Szymanski, M. T., Petric, M. Am. J. Dis. Child. 1977, 131,
733.
PREVENTION OF IRON LOADING IN
TRANSFUSION-DEPENDENT THALASSÆMIA
M. J. PIPPARD* E. A. LETSKY†
S. T. CALLENDER* D. J. WEATHERALL*
*Nuffield Department of Clinical Medicine, Radcliffe
Infirmary, Oxford, and †Hospital for Sick Children, Great
Ormond Street, London
Summary Urinary iron excretion after single intra-
muscular (i.m.) bolus injections or 12 h
subcutaneous (s.c.) infusions of desferrioxamine (D.F.)
was determined in sixteen homozygous &bgr;-thalassæmia
patients whose ages ranged from 10 months to 23 years.
At all ages the s.c. infusions resulted in greater iron loss
than identical i.m. doses. With doses of 0·5-1 g of D.F.
as s.c. infusions eight out of nine children aged less than
6 years with a total transfusion iron load of less than
10 g excreted sufficient iron to achieve iron balance.
These results suggest that iron loading in transfusion

dependent thalassæmics may be preventable by the use
of an appropriate chelation regimen started early in life.
Introduction
IN thalassæmic children dependent upon regular
blood-transfusions studies of liver-biopsy specimens sug-
gest that tissue damage from excess iron starts early in
infancy. Although daily intramuscular (i.m.) injections
of the iron-chelating agent desferrioxamine (D.F.) may
slow up the progression of pathological changes,2 this
regimen does not result in urinary iron excretion which
is sufficiently high to keep pace with transfusion iron
loading until the age of about 6 years, by which time at
least 10 g of excess iron has accumulated.3 Subc.u-
taneous (s.c.) infusions of D.F. produced greater iron
excretion than bolus doses, especially in ascorbate-
replete subjects.4-7 We studied the effectiveness of this
new approach and the possibility of preventing toxic
iron accumulation in thalassaemic children under the age
of 6.
Patients and Methods
Sixteen transfusion-dependent thalasssemic patients were in-
vestigated (see accompanying table). The patients fell into two
age-groups. Nine children were aged less than 6 years and all
had received less than 10 g.of iron through blood-transfusion.
They had been treated with similar transfusion regimens with
haemoglobin values maintained above 9-0 g/dl. None had
received previous chelation therapy or required splenectomy.
Seven patients aged more than 6 years formed a less uniform
group, though all had heavy iron loads (11-69 g). Past trans-
fusion policies had varied, though all were currently on a pro-
gramme similar to that of the younger children. Five had
received i.m. D.F. for variable periods. All had needed splenec-
tomy because of an increase in blood-transfusion requirements.
For each patient daily iron loading from blood-transfusions
over the previous 6 months was calculated (1 unit of blood
being equivalent to 200 mg of iron).
Urinary iron excretion was determined after i.m. bolus and
12 h s.c. infusions of the same D.F. dose (05gin all but one
PATIENT DETAILS WITH URINARY IRON EXCRETION AFTER I.M.
1179
u J IU :J
I
DoiLy transfusion iron load (mg)
Fig. 1-Urinary iron excretion in response to bolus i.m. injection
(e) and 12 h s.c. infusion (0) of equal doses of D.F. (0.5 or 1
g), in patients aged less than 6.
Dotted line represents iron balance between daily transfusion iron
load and urinary iron excretion.
of the patients aged less than 6 years and 1 g in all but two
of the patients aged more than 6 years). In addition dose-
response curves to the larger doses of D.F. which can be given
as 12 h s.c. infusions were plotted. Infusions were given at the
rate of 1-2 ml per hour into the subcutaneous tissues of the
anterior abdominal wall by means of a 23-gauge butterfly
needle connected by a narrow-bore tube to a Handley clock-
work infusion pump.’ 24 h urinary iron excretion after each
i.m. bolus injection or 12 h s.c. infusion was measured with a
Perkin-Elmer absorption spectrophotometer.8
Because ascorbic-acid saturation increases iron excretion
after administration of D.F.s-7,9 and recent blood-transfusion
reduces it,10 all studies were carried out after ascorbate reple-
tion and as close as possible to the midpoint between trans-
fusions.
Results
Comparison of Bolus with Infusion
When the calculated daily transfusion iron load was
compared with the urinary iron excretion obtained with
equal doses of D.F. given by the two routes none of the
younger children achieved iron balance with the bolus
BOLUS AND 12 H S.C. INFUSION OF D.F. DOSE SHOWN
 1180
injection, whereas most of this group obtained balance
with the s.c. infusion (table, fig. 1).
In contrast, all except one of the older children (case
11) showed a net iron excretion even with the bolus in-
jections.
However, the same dose given as an infusion was
much moreeffective (fig. 2). The mean percentage in-
crease in iron excretion when infusion was compared
with bolus administration was, greater in the younger
group (97%) than in the older (58 %) group.
Dose Response with Subcutaneous Infusion
In the younger age-group iron excretion with increas-
ing doses of D.F. tended to level off above a dose of 1 g.
In only one patient (case 8) was there a linear response
(fig. 3). With doses of 0.5-1 g, all but one (case 2) of the
younger children achieved iron balance with respect to
daily transfusion iron load, and many showed a substan-
tial net iron excretion., In the older patients 4) there (fig.
was less of a tendency for the curve to flatten at higher
doses, and in two patients .(cases 13 and 16) a linear re-
sponse was maintained up to a 4 g dose. In these two
patients iron excretion equivalent to 0.75 of a unit of
blood was possible with a single 12 h infusion at the
highest dose.
Iron excretion increased with age, and with a stan-
dard infusion of 2 g D.F. in 12 h there is a highly signifi-
cant correlation with both age (r 0.93, P<0.001) and
=
the amount of previous transfusion iron load (r 0.87,
P<0.001).
Discussion
Previous studies with i.m. D.F. have suggested that no
significantiron excretion can be achieved in thalassasmic
Daily transfusion iron load (mg)
Fig. 2-Urinary iron excretion in response to bolus i.m. injection
(e) and 12 h s.c. infusion (0) of equal doses of D.F. (0.5 or 1
g) in patients aged more than 6.
Dotted line represents iron balance between daily transfusion iron
load and urinary iron excretion.
J
D.Flg) in 12h s.c infusion
Fig. 3-Urinary iron excretion in patients aged less than 6
given increasing doses of D.F. as 12 h s.c. infusions.
D.F.(g) in 12h s.c. infusion
Fig. 4-Urinary iron excretion in patients aged more than 6
years given increasing doses of D.F. as 12 h s.c. infusions.
children until the age of about 6 years, by which time
a considerable and toxic iron load has accumulated. The
present studies havecorifirmed that bolus injections of
D.F. have little effect in the younger age-group but have
shown that s.c. infusions can be used to obtain a nega-
tive iron balance, even in young children with low iron
load.
Dose-response curves indicate that in most cases there
is no need to give more than 0.5-1 g by s.c. infusion to
=
achieve maximum iron excretion in the younger children
whereas in older patients larger doses usually produce a
greater effect. The dose of D.F. should be tailored to the
needs of the individual patient.’ In most cases the opti-
mum dose will be in the range 50-100 mg/kg. This is
within the range known to produce cataracts in dogs,"
and in long-term use careful monitoring of the eyes is
essential.
Iron-balance studies with both bolus injections’2 and
12 h s.c. infusions’&deg; showed that about a third of the
total iron excretion occurs in the faeces. The measure-
ment of urinary iron excretion alone therefore underesti-
mates total iron excretion. With regular use of s.c. D.F.
infusions for up to 2 years in Boston5 and Oxford’ there
has been no evidence of tachyphylaxis in children aged
3 years and over. Combined with the present studies
these observations indicate that in young children the
build-up of iron and eventual cardiac, hepatic, and en-
docrine damage may be preventable. Further work is
needed on alternative ways of getting a slow release of
D.F. into these children,
since daily use of the current in-
fusions, even if only during sleeping hours, is cumber-
some and would be impractical in those parts of the
world where thalasssemia is common.
We are grateful to Mrs J. Bull for her excellent technical assistance.
This work was supported by grants from the Medical Research Coun-
cil, London, and the Oxfordshire Area Health Authority (T).
Correspondence should be addressed to M. J. P., Nuffield Depart-
ment of Clinical Medicine, Radcliffe Infirmary, Oxford OX2 6HF.
REFERENCES
1. Iancu, T. C., Neustein, H. B., Landing, B. H. in Iron Metabolism; p. 293.
Ciba Fdn Symp. 51, Amsterdam, 1977.
2. Barry, M., D. M., Letsky, E. A., Risdon, R. A. Br. med. J. 1974, ii,
16.
Flynn,
3. Modell, C. B, Beck, J. Ann. N.Y. Acad. Sci. 1974, 232, 201.
years
 1181

COLONIC FUNCTION IN PATIENTS WITH toAnalyzer’ after charring with nitric acid), and p.E.G.5 and
DIVERTICULAR DISEASE Cr 203were estimated in fifty patients. In forty-one patients
colonic motility was tested. Colonic motility was studied with
M. A. EASTWOOD A. N. SMITH open-ended tubes by means of which both basal measurements
and response to food were measured in the distal colon and rec-
W. G. BRYDON J. PRITCHARD tum. Measurements were taken 5, 15, and 25 cm from the anal
verge. The amplitude and frequency of the waves were calcu-
lated to give the motility index.7
Wolfson Laboratories, Gastrointestinal Unit, Departments of
Medicine and Clinical Surgery, University of Edinburgh, and
Department of Clinical Chemistry, Western General Hospital, Results
Edinburgh The mean stool weight was 98 g/24 h (range 20-190

Summary Diverticular disease is thought to be as- g/24 h. The transintestinal transit-times ranged from 24
to 190 h (mean 53 h). In the thirty subjects who took
sociated with prolonged intestinal tran-
sit-time, a reduced stool weight, and increased intraco-
P.E.G./Cr203 capsules, the P.E.G./chromium ratio ranged
from 0 - 2 to 1-7 (mean of 0-83). Faecal-fat excretion var-
lonic pressure. Sixty patients with diverticular disease
did not regularly show these features. Variation in col-
onic function was considerable in these patients and was
similar to that in the general population from which the
patients were recruited. Constipation may be a compli-
cation of diverticular disease and not necessarily part of
its &aelig;tiology.

Introduction
THE management of diverticular disease has changed
radically. The rationale for the introduction of dietary
fibre into the treatment regimen is based on observations
that diverticular disease is characterised by a low stool
weight, prolonged intestinal transit-time, and raised in-
tracolonic pressure, all of which are thought to be the
result of a reduced intake of dietary fibre. However,
since it is not known whether these changes are always
present, we investigated colonic function in sixty pa-
tients with diverticular disease.

Patients and Methods

Sixty patients with diverticular disease were investigated,
but not all in the same detail. All the patients had been at
home on their usual diet and were investigated as outpatients
after consenting to inclusion in the study. The mean age was
Motility index of rectosigmoid region in 41 patients with diver-
60 years, range 32-84 years. Diagnosis was based on a bar- ticular disease.
ium-enema examination preceded by sigmoidoscopy and a
clinical examination. Symptoms included constipation,
diarrhoea, and pain. In some the disease was asymptomatic and
diagnosis was the result of a chance finding. Sixty patients ied from 1 to 19 mmol/24 h (mean 8.0 mmol/24 h).
recorded their transintestinal transit-time by means of barium- Fxcal bile-acid excretion ranged from 0.11 to 1.4
impregnated markers;2 such markers not only give a crude esti- mmol/24 h (mean 0.5mmol/24 h). The mean basal
mate of transit-time but indicate the accuracy of stool collec-
motility index in forty-one patients was 800 and the
tion. Thirty patients took polyethylene glycol/chromium
post-food motility index was 2200 (range 400-6800) (see
sesquioxide (P.E.G./Cr203) capsules to demonstrate gastrointes- accompanying figure).
tinal streaming.3 The stool was collected and stored at &mdash;20&deg;C,
pooled, and freeze dried. Faecal wet weight, bile acids,4 fat,
Discussion
electrolytes, sodium, potassium, calcium, and magnesium (’Au-
None of these patients had been on any kind of treat-
ment they were studied, and they may be
before
4. Hussein, M A M , Flynn, D. M., Green, N., Hussein, S., Hoffbrand, A. V.
regarded as being typical of patients coming to an out-
Lancet, 1976, ii, 1278. patient department. Our results raise some doubts about
5. Propper, R D., Cooper, B., Rufo, R. R., Neinhuis, A. W , Anderson, W. F., the overall importance of the "features" of diverticular
Bunn, H. F, Rosenthal, A., Nathan, D. G. New EnglJ Med. 1977, 297,
418. disease-i.e., low stool weight, prolonged transit-time,
6. Hussein, M. A M., Flynn, D. M., Green, N., Hoffbrand, A. V. Lancet, 1977, and raised intraluminal pressure-which were formerly
i,977.
7 Pippard, M. J., Callender, S. T., Weatherall, D. J Clin. Sci. molec. Med. thought to be characteristic of this disease. The stool
1978, 54, 99.
8. Zettner, A., Mansbach, L. Am. J. clin. Path. 1965, 44, 517.
weight and transit-time in our patients with untreated
9. Wapnick, A. A., Lynch, S. R., Charlton, R. W., Seftel, H. C., Bothwell,
diverticular disease were in the same range as those in
T. H Br J. H&aelig;mat. 1969, 17, 563 a normal population-80 g/24 h (range 20-280 g) and
10 Pippard, M J., Callender, S. T. Unpublished. 72 h (range 24-168 h), respectively--derived from the
11. Ciba Pharmaceutical Company. Clin. Pharmac. Ther. 1969, 10, 595.
12 Cumming, R L. C., Millar, J. A., Smith, J. A., Goldberg, A. Br. J. H&aelig;mat. same part of Edinburgh and aged 16-80 (unpublished).
1969, 17, 257. The motility index in the diverticular-disease patients