606
difference. When the cases of possible nephrotoxicity
and those in which placebo was substituted for methicil-
lin included, the results are still statistically signifi-
are
cant. Thus, we have shown in a prospective, controlled,
double-blind study that the combination of cephalothin
plus an aminoglycoside is more nephrotoxic than the
combination of methicillin plus an aminoglycoside.
In animals, cephalothin does not potentiate aminogly-
coside nephrotoxicity but in some cases actually seems to
protect against it.11-14 The efficacy and toxicity of
cephalothin plus an aminoglycoside have been compared
with a variety of other combinations containing an
aminoglycoside. The results of some of these studies sup-
port increased nephrotoxicitv with the cephalothin-
aminoglycoside combination7 8 and others do not. 9 10
The disparity among these previous studies and their
failure to support conclusively an interaction between
cephalothin and an aminoglycoside may be due to a var-
iety of factors. For example, the applicability to man of
studies done in animals has not been demonstrated.
None of the clinical studies were double-blind, so that
bias might have influenced the evaluation. Furthermore,
in most cases, nephrotoxicity was not clearly defined and
serum-aminoglycoside levels were not measured. Conse-
quently, serum-aminoglycoside levels may have differed
between the groups. Since nephrotoxicity is believed to
be correlated with high serum-levels of aminoglycosides,
high levels in the cephalothin groups may have
accounted for the results. Finally, the two studies sup-
porting nephrotoxicity were done in patients with
cancer, whereas the studies which failed to demonstrate
the interaction were done in patients with general medi-
cal problems. Consequently, patient selection may have
influenced the results.
Our study was double-blind, and therefore bias could
not have influenced the results. In addition, we have
stated our definition of nephrotoxicity, measured
aminoglycoside levels, and treated a general medical
population. These characteristics give our results added
validity. We believe our work substantiates a nephro-
toxic interaction between cephalothin and aminoglyco-
sides in man.
We are indebted to Dr David Mellits for his statistical advice, Miss
Judy Hyman for nursing assistance, Mrs Marguerite Sonneborn and
Mrs Jean K. Linz for technical assistance, Mrs Evelyn Fleckenstein
and Ms Gail Otis for assistance, and members of the Osler medical
house-staff of the Johns Hopkins Hospital for their cooperation. This
study was supported by the Eli Lilly Company, Indianapolis, Indiana.
Requests for
reprints should be addressed to C.R.S., Division of
Climcal Pharmacology, Johns Hopkins Hospital, 600 N Wolfe Street,
Maryland 21205, U.S.A.
REFERENCES
1. Opitz, A., Herrmann, I. V., Herrath, D., Herrath, D.v., Schoefer, K.
Medsche Welt, Berl. 1971, 11, 434.
2. Bobrow, S., Jaffe, E., Young, R.J.Am.med.Ass. 1972, 222, 1546.
3. Kleinknecht, D., Ganeval, D., Droz, D. Lancet, 1973, 1,1129.
4. Fillastre, J. P., Laumonier, R., Humbert, G., Dubois, D., Metayer, J., Del-
pech, A., Leroy, J., Robert, M. Br. med J. 1973, ii, 396.
5. Cabanillas, R., Burgos, R. C., Rodriguez, R. C., Baldizón, C. Archs intern.
Med. 1975, 135, 850.
6. Plager, J. R. Cancer, 1976, 37, 1937.
7. Klastersky, J., Heusgens, C., Debusscher, L. Antimicrob. Ag. Chemother.
1975, 7, 640.
8. Schimpff, S. C., Gaya, H., Klastersky, J., et al. J. infect. Dis. 1978, 137, 14.
9. Fanning, W. L., Gump, D., Jick, H. Antimicrob. Ag. Chemother. 1976, 10,
80.
10. Hansen, M. M., Kaaber, K. Acta med. scand. 1977, 201, 463.
References continued at foot of next column
Preliminary Communications
ABNORMAL RECTAL IMMUNOGLOBULIN
PATTERN IN HIRSCHSPRUNG’S DISEASE
THOMAS C. HALPIN, JR RONALD P. GREGOIRE
ROBERT J. IZANT, JR
Departments of Pediatrics and Surgery, Case Western Reserve
University School of Medicine, and the Rainbow Babies and
Childrens Hospital, Cleveland, Ohio, U.S.A.
Summary immunoglobulin content of rectal
The
biopsy tissue and secretions from twelve
neonates in whom Hirschsprung’s disease was suspected
was determined by an organ-culture technique and
radioimmunoassay. The immunoreactive IgG content of
explanted rectal tissue and its secretions in those six
children who proved to have Hirschsprung’s disease was
much higher than in those with other types of obstruc-
tive lower-intestinal disease. Increased amounts of IgG
may represent maternally derived antibody associated
with neonatal gut neural elements. This seems to be the
first report of an immunological abnormality in Hirschs-
prung’s disease.
INTRODUCTION
HIRSCHSPRUNG’S disease has well recognised clinical,
radiographic, and pathological features.’ Numerous
reports have described the morphological and histoche-
mical abnormalities present in aganglionic segments,
and diagnosis has classically depended on the absence of
the ganglion cells of the myenteric and submucosal plex-
uses. 23 However, no definite mechanism has been sug-
gested to explain this abnormality.
Intestinal immunoglobulins were assayed in rectal-
biopsy specimens from newborn infants in whom
Hirschsprung’s disease was suspected. Rectal tissue was
maintained in an organ-culture system for 24 h and then
class-specific immunoglobulins in tissue homogenates
and secretions were measured by a single-phase antibody
radioimmunoassay (R.I.A.). We found an unusual rectal
immunoglobulin pattern in Hirschsprung’s disease.
PATIENTS AND METHODS
Patients
Twelve neonates presenting within the first week of life with
symptoms and/or physical findings suggestive of intestinal
obstruction underwent rectal biopsy as part of a diagnostic
evaluation. Proctosigmoidoscopy was performed in all children
DR WADE AND OTHERS: REFERENCES—continued
11. Harnson, W. O., Silverblatt, F. J., Turck, M. Antimicrob. Ag. Chemother.
1975, 8, 209.
12. Dellinger, P., Murphy, T., Pinn, V., Barza, M., Weinstein,
L. ibid. 1976, 9,
172.
13. Luft, F. C., Patel, V., Yunm, M. N., Yum, M. N., Kleit, S. A. ibid. p. 831.
14. Bloxham, D. D., Meliere, C. C., Thompson, W. L. Clin. Res. 1978, 26,
286A.
15. Smith, C. R., Baughman, K. L., Edwards, C. Q., Rogers, J. F., Lietman,
P. S.New Engl.J.Med. 1977, 296, 349.
16. Chan, R. A., Benner, E. J., Hoeprich, P. D. Ann. intern. Med. 1972, 76,
773.
17. Lietman, P. S., Lipsky, J. J., Johannes, R. S. Abstracts of American Society
for Microbiology, 72nd Annual Meeting, Philadelphia, April 1972, p. 121.
Washington D.C., 1972.
18. Smith, D. H., Van Otto, B., Smith, A. L. New Engl. J. Med. 1972, 286, 583.

except those with suspected Hirschsprung’s disease. Rectal-
biopsy specimens were obtained either with the ’Quinton’ mul-
tipurpose suction tube with a 1 - 5 mm biopsy port or as a full-
thickness biopsy. A portion of the surgical sample was
submitted for organ culture to determine total and secreted
immunoglobulins. The morphology of the remaining specimen
was investigated by the Department of Pathology and one of
us (T.C.H.).
Rectal Organ Culture
After transport in cold Hanks’ balanced salt solution rectal
explants were placed with the villous surface uppermost on a
stainless steel mesh in the centre well of an organ-culture dish
according to the method of Eastwood and Trier.4 Centre-well
secretions from explants were obtained after a 24 h incubation
period at 37°C in 95% oxygen and 5% carbon dioxide. The
volume was recorded and the sample frozen at -70°C.
Explants were transferred to a lysis buffer (0-5mol/I phos-
phate buffer, pH=7.4) and homogenised. A sample was taken
to determine total biopsy protein.5 Nuclear and cytoplasmic
debris was removed by centrifugation at 20 000 g for 30 min
at 4°C; supranatants were stored at -70°C.
Radioimmunoassay
The IgA, IgM, and IgG content of explant homogenates and
secretions was determined by a single-antibody phase radioim-
munoassay.6 Briefly, the immunoglobulin fraction of heavy-
chain-specific human antisera was coupled to cyanogen-bro-
mide-activated ’Sephadex G-25’ (particle size 10-40 pLm) by a
modification of the method of March et al. for the preparation
of immunoabsorbent.7 Purified human IgM, IgG, and IgA
antigens were radiolabelled (12s1) by lactoperoxidase-catalysed
radioiodination.8 After determining optimum concentrations
of 125 I-immunoglobulin and immunoabsorbent, the tubes were
incubated with 20 and 50 ul of explant and secretion superna-
tants at room temperature for 16-18 h with continuous agi-
tation. After washing and centrifugation, radioactivity was
counted in an automatic gamma counter. Results were
expressed as g of immunoglobulin per mg of intestinal tissue
pxotein. Amounts of secreted immunoglobulins were calculated
by dividing total )JLg per period of secretion by the tissue-pro-
tein concentration.
RESULTS
Comparison of rectal IgG, IgM, and IgA measure-
ments in the patient groups showed that amounts of IgG
in both secretions and tissue samples from patients with
Hirschsprung’s disease were significantly higher than in
RECTAL IMMUNOGLOBULINS AFTER 18-HOUR INCUBATION AT
IN 95% OXYGEN 5% CARBON DIOXIDE ATMOSPHERE
607
children with other neonatal intestinal obstruction syn-
dromes (P<0.001) (see accompanying table). Also
amounts of immunoreactive IgA in secretion were sig-
nificantly greater (P<0.015) in neonates with Hirsch-
sprung’s disease than in those with intestinal obstruc-
tion. Total IgM levels were normal for age. Intestinal
tissue and secretion immunoglobulin measurements in
eleven older children with functional bowel disease are
included for comparison (see table). Pathology of the
rectal specimens in children with Hirschsprung’s disease
revealed the absence of ganglion cells, and no evidence
of inflammation or mucosal disease was found. Plasma-
cells were seen in normal numbers in all patients. Clini-
cally, the other neonatal obstruction syndromes were
meconium plug syndrome (three cases), left colon syn-
drome (one case), pseudo-obstruction syndrome (one
case), and pelvic abscess (one case).
DISCUSSION
The pathogenesis of Hirschsprung’s disease is still un-
known. Early inhibition of neuroblast migration
between the 8th and 12th weeks of gestation and/or the
destruction of migrating neurons of the Auerbach and
Meissner myenteric plexus may be responsible for the ,
observed intestinal lesion. However, the apparent in-
crease in IgG in rectal tissue and secretions in neonates
with Hirschsprung’s disease is extremely interesting.
The possibility that IgG class-specific maternal anti-
bodies directed against fetal ganglion cells may partici-
pate (with or without other host factors) in an immuno-
logical reaction resulting in loss of normal colonic
innervation has not been considered. Such modifying
factors may include bacterial endotoxin as an adjuvant
and/or food macromolecular antigens.
Although gastrointestinal tissue in neonates can pro-
duce immunoglobulins it also contains systemically de-
rived immunoglobulins. During the first month of life,
placentally transferred maternal IgG is the major im-
munoglobulin. As the child’s own intestinal immune sys-
tem is stimulated by food and bacterial and environmen-
tal antigens, intrinsic immunoglobulin levels rise both in
the systemic and secretory immune system.9 The exact
timing of intestinal immunoglobulin production is not
known but secretory antibodies (primarily S-IgA) pro-
duced in response to antigen stimulation begin to appear
within two weeks of birth in rats.l0 11 These levels rise
370c rapidly in intestinal secretions and seem to protect the
human neonate from enteric infections.
We do not know how rectal IgG becomes increased in
Hirschsprung’s disease. Significant intestinal biosyn-
thesis of IgG is not thought to occur before the 4th week
of life. Inflammation secondarv to colonic stasis is an un-
likely cause since inflammatory cells are not found in the
lamina propria of patients with Hirschsprung’s disease.
In addition, the low levels of IgG in rectal specimens in
other obstruction syndromes argues against colonic
stasis as an a priori cause of increased IgG. Maternally
derived immune-reactive proteins have been identified in
other neonatal diseases. Haemolytic diseases of neonates
caused by Rh isoimmunisation (including other
Coombs-positive blood-group incompatibilities), hyper-
thyroidism secondary to transplacental passage of
the 7S globulin (long-acting thyroid stimulator), im-
mune thrombocytopenia with maternal antibodies dir-
 608

ected against platelets from neonates, and discoid lupus

secondary to maternal systemic lupus erythematosis are
all examples of maternal IgG-associated neonatal dis-
ease.12-15
Our preliminary data indicate that there may be an
immunological component in Hirschsprung’s disease
and this may explain the increased familial incidence.
This possibility should be intensively investigated.
We thank Mrs J. Zollinger for valuable technical assistance and our
paediatrician colleagues, particularly Dr R. E. Behrman and Dr S. Pol-
mar, for their helpful discussions. The Rainbow Foundation gave
financial support.

Requests for reprints should be addressed to T.C.H., Jr., Case Wes-

ternReserve University School of Medicine, Department of Pediatrics,
2101 Adelbert Road, Cleveland, Ohio 44106, U.S.A.

REFERENCES

1. W. Meier Ruge. Curr. Topics Path. 1974, 59, 131.

2. Baumgarten, H. G., Holstein, A. F., Stelzher, F. Virchows Arch. path. Anat.
Physiol, 1973, 358, 113.
3. Howard, E. R. Am. Surg. 1973, 39, 602.
4. Eastwood, G. L., Trier, J. S. Gastroenterology, 1973, 64, 375.
5. Lowry, O. H., Rosenbrough, N. J., Farr, A. L., Randall, R. J.J. biol. Chem.
1951, 193, 265.
6. Wide, L.Acta endocr., Copenh. 1969, 142, suppl. p. 207.
7. March, S. C., Parkh, I., Ovatrecasas, P. Analyt. Biochem. 1974, 60, 149.
8. Marchalonis, J. J. Biochem.J. 1969, 113, 299.
9. Head, J. R. Semin. Perinat. 1977, i, 2.
10. Nagura, Hiroshi, Nakane, P. K., Brown, W. R.J. Immun., 1978, 120, 4
11. Kagnoff, M. F. ibid. 1977, 118, 3.
12. Zipursky, A. in Hematology of Infancy and Childhood (edited by D. G.
Nathan and F. A. Oski); p. 137. Philadelphia, 1974.
13. Riopel, D. A., Mullins, C. E. Pediatrics, 1972, 50, 140.
14. Pearson, H. A., Shulman, N. R., Maden, V. J., Cone, T. E. Blood, 1964, 23,
154.
15. Reed, W. B., May, S. B., Tuffanelh, D. L. Archs Derm. 1967, 96, 64.
Fig. 1--Cerebropontine-angle meningioma. Upper figure: stan-
dard gamma-camera brain scans. Lower figure: on the left,
emission tomographic brain scan (Cleon); on the right,
transmission tomographic brain scan (EMI).
Tumour well seen by all three imaging modalities. Emission tomo-
EMISSION COMPUTERISED TOMOGRAPHY graphic brain scan shows surprising detail of normal structures of the
anterior, middle, and posterior fossae, as well as the abnormal struc-
A New Diagnostic Imaging Technique ture (tumour).

P. J. ELL P. H. JARRITT
J. DEACON N. J. G. BROWN
E. S. WILLIAMS
Middlesex Hospital Medical School, London WI
EMISSION computerised axial tomography (E.C.A.T.) is
a new non-invasive imaging technique. It was first applied
to the brain and has now been applied to the body.2 As
with transmission computerised tomography (EMI type
of imaging), a computer is used to reconstruct a two-
dimensional image. With E.C.A.T. radiation emitted from
the organ and recorded by scintillation detectors is used
for image processing. This contrasts with EMI type of
imaging, in which a beam of X-rays shines through the
body and absorption coefficients are used for image
reconstruction. We present here original data from an
emission tomographic imager installed in March, 1978,
at the Middlesex Hospital.
from the centre (aperture 25 cm). Each of the 12 detectors
consists of a sodium-iodide crystal, a focused collimator, a pho-
tomultiplier tube, and associated electronics. Up to 8 slices can
be scanned automatically. These are chosen by the operator at
multiples of 3 mm apart. The final image, once reconstructed,
reflects a cross-sectional distribution of the radiopharmaceuti-
cal within the organ. Polaroid or transparent X-ray-film hard-
copy output is available.
In the initial clinical evaluation of the emission tomographic
scanner, over 80 patients were investigated with standard
99mtechnetium-labelled radiopharmaceuticals. The scanner was
operated with window settings for 99mtechnetium (130-170
Kev), a scan time of four minutes per slice and a slice spacing
of 1-25 cm. In the case of brain imaging with 99ritechnetium
pertechnetate, a typical background cut-off of 25% was
chosen. Whenever possible, a comparison was made with other
non-invasive imaging techniques, such as EMI imaging and
conventional analogue gamma-camera imaging.

MATERIALS AND METHODS

The scanner (Cleon-710 transverse-section brain imager) is RESULTS

designed as a brain imager. It has a gantry assembly which can
be tilted to an angle of &plusmn;20&deg; to the vertical, a computer, and
an operator console. It scans the brain with 12 focused radia-
tion detectors equally spaced around the centre. The detectors
move in pairs in a rectilinear sequence, such that when one
pair of detectors is scanning and incrementing towards the
centre, the adjacent pairs are scanning and incrementing away
In a series of 40 brain scans, no false positives were
encountered. From a total of 15 lesions diagnosed on the
EMI scanner, 14 were seen by the emission scanner and
12 were detected by conventional gamma-camera imag-
ing. 1 false-negative case (cyst) was seen by the
transmission X-ray scanner but not by the emission