Intravenous immunoglobulin therapy for severe Clostridium difficile colitis

Abstract

Many individuals have serum antibodies against Clostridium difficile toxins. Those with an impaired antitoxin response may be susceptible to recurrent, prolonged, or severe C difficile diarrhoea and colitis.
To examine whether treatment with intravenous immunoglobulin might be effective in patients with severe pseudomembranous colitis unresponsive to standard antimicrobial therapy.
Two patients with pseudomembranous colitis not responding to metronidazole and vancomycin were given normal pooled human immunoglobulin intravenously (200-300 mg/kg).
Antibodies against C difficile toxins were measured in nine immunoglobulin preparations by ELISA and by cytotoxin neutralisation assay.
Both patients responded quickly as shown by resolution of diarrhoea, abdominal tenderness, and distension. All immunoglobulin preparations tested contained IgG against C difficile toxins A and B by ELISA and neutralised the cytotoxic activity of C difficile toxins in vitro at IgG concentrations of 0.4-1.6 mg/ml.
Passive immunotherapy with intravenous immunoglobulin may be a useful addition to antibiotic therapy for severe, refractory C difficile colitis. IgG antitoxin is present in standard immunoglobulin preparations and C difficile toxin neutralising activity is evident at IgG concentrations which are readily achieved in the serum by intravenous immunoglobulin administration.

Full text

Intravenous immunoglobulin therapy for severe
Clostridium diYcile colitis
J Salcedo, S Keates, C Pothoulakis, M Warny, I Castagliuolo, J T LaMont, C P Kelly
Abstract
Background—Many individuals have
serum antibodies against Clostridium dif-
ficile toxins. Those with an impaired anti-
toxin response may be susceptible to
recurrent, prolonged, or severe CdiYcile
diarrhoea and colitis.
Aims—To examine whether treatment
with intravenous immunoglobulin might
be eVective in patients with severe pseu-
domembranous colitis unresponsive to
standard antimicrobial therapy.
Patients—Two patients with pseudomem-
branous colitis not responding to metroni-
dazole and vancomycin were given normal
pooled human immunoglobulin intrave-
nously (200–300 mg/kg).
Methods—Antibodies against CdiYcile
toxins were measured in nine immu-
noglobulin preparations by ELISA and by
cytotoxin neutralisation assay.
Results—Both patients responded quickly
as shown by resolution of diarrhoea,
abdominal tenderness, and distension. All
immunoglobulin preparations tested con-
tained IgG against CdiYcile toxins A and
B by ELISA and neutralised the cytotoxic
activity of CdiYcile toxins in vitro at IgG
concentrations of 0.4–1.6 mg/ml.
Conclusion—Passive immunotherapy with
intravenous immunoglobulin may be a
useful addition to antibiotic therapy for
severe, refractory CdiYcile colitis. IgG
antitoxin is present in standard immu-
noglobulin preparations and CdiYcile
toxin neutralising activity is evident at IgG
concentrations which are readily achieved
in the serum by intravenous immu-
noglobulin administration.
(Gut 1997; 41: 366–370)
Keywords: Clostridium diYcile; toxin; diarrhoea; IgG;
immunotherapy; antibiotic
Clostridium diYcile antibiotic associated colitis
is an important cause of morbidity in hospital
and nursing home patients.
1
As many as 26% of
hospital patients are colonised by CdiYcile and
up to one third of these develop diarrhoea.
2
There is a wide spectrum of host response to C
diYcile infection ranging from asymptomatic
carriage to fulminant colitis with toxic
megacolon.
1
CdiYcile colitis is mediated by two
large protein exotoxins released by pathogenic
strains of the bacterium.
1 3–7
ToxinAisa
308 kDa protein which is both a cytotoxin and
a potent inflammatory enterotoxin.
8–10
Toxin B,
a 280 kDa protein, is a more potent cytotoxin
than toxin A but is not enterotoxic for rodent
intestine.
711
However, toxin B may be even
more harmful to human colon than toxin A.
6
Both toxins share the same intracellular mech-
anism of cytotoxicity. They act as enzymes to
glucosylate a threonine residue on small GTP
binding rho proteins.
12 13
This leads to the dis-
aggregation of actin filaments, collapse of the
cytoskeleton, and cell rounding.
The presence or absence of an adequate
antibody response to CdiYcile toxins may play
an important role in determining the severity of
diarrhoea and colitis.
14
Serum antibodies to
toxins A and B are evident in two thirds of
healthy adults.
15–17
Patients with low antitoxin
antibody levels are reported to have more
severe, more prolonged, or recurrent CdiYcile
diarrhoea whereas asymptomatic carriers have
higher antitoxin levels.
18–23
We previously re-
ported that children with recurrent CdiYcile
diarrhoea had low levels of serum IgG against
CdiYcile toxin A.
21
When these children were
treated with intravenous immunoglobulin their
serum antitoxin levels increased and their diar-
rhoea resolved. We now report the use of
intravenous immunoglobulin therapy in two
adults with severe pseudomembranous colitis
which failed to respond to standard antibiotic
treatment with metronidazole and vancomy-
cin. We also demonstrate that normal human
pooled immunoglobulin contains antibodies
against CdiYcile toxins A and B which can
neutralise the cytotoxic eVects of these toxins.
Methods
MEASUREMENT OF ANTI-C DIFFICILE IgGIN
IMMUNOGLOBULIN PREPARATIONS
Nine human immunoglobulin preparations
intended for intravenous administration were
studied. Three batches of immunoglobulin
were obtained from each of the following pro-
ducers: Alpha Therapeutic Corporation (Los
Angeles, California, USA), Armour Pharma-
ceutical Company (Kankakee, Illinois, USA),
and Baxter Healthcare Corporation (Glendale,
California, USA). All were highly purified
preparations of intact unmodified IgG isolated
from large pools of human plasma by cold
alcohol fractionation.
Human IgG levels to CdiYcile antigens were
measured by enzyme linked immunosorbent
assay (ELISA) as previously described.
15 21 24
IgG directed against highly purified CdiYcile
toxins A and B and against a culture filtrate of
toxigenic CdiYcile (strain VPI 10463) were
measured separately. The CdiYcile culture fil-
trate contains toxins A and B as well as
non-toxin CdiYcile antigens. ELISA results are
Gut 1997; 41: 366–370366
Section of
Gastroenterology,
Boston University
School of Medicine,
Boston,
Massachusetts, USA
J Salcedo
Gastroenterology
Division, Beth Israel
Deaconess Medical
Center, Harvard
Medical School,
Boston,
Massachusetts, USA
MWarny
S Keates
C Pothoulakis
I Castagliuolo
J T LaMont
C P Kelly
Correspondence to:
Dr C P Kelly, Dana 601,
Gastroenterology, Beth Israel
Deaconess Medical Center,
330 Brookline Avenue,
Boston, MA 02215, USA.
Accepted for publication
2 May 1997

expressed as optical density readings at
450 nm.
15 21 24
MEASUREMENT OF C DIFFICILE TOXIN
NEUTRALISING ACTIVITY IN IMMUNOGLOBULIN
PREPARATIONS
Cytotoxicity was determined by rounding of
fibroblasts (R9AB, American Type Culture
Collection, Rockville, Maryland, USA) in
monolayer culture after exposure to CdiYcile
toxins.
52425
The minimum 50% cytotoxic dose
for each toxin preparation, defined as the mini-
mum dose resulting in 50% cell rounding at 24
hours, was 0.1 ng/ml for toxin A, 0.003 ng/ml
for toxin B, and 0.5 ng/ml for culture filtrate in
these experiments. Inhibition of cytotoxicity
was quantified by adding serial twofold dilu-
tions of the immunoglobulin preparations to
four times the minimum 50% cytotoxic dose of
each toxin preparation. After 20 minutes the
toxin/immunoglobulin mixture was added to
fibroblast monolayer cultures and cell round-
ing was assessed after 24 hours. Controls
included human serum albumin diluted to the
same protein concentration as the IgG and
serum from a healthy volunteer who lacked
specific antibodies against CdiYcile toxin A or
toxin B as determined by ELISA.
15 23
Results
are expressed as the lowest concentration of
human IgG required to prevent rounding of
50% of the fibroblasts.
24
Case reports and Results
PATIENT 1
A 63 year old woman developed diarrhoea,
cramping abdominal pain, and abdominal dis-
tension five days after laparotomy for staging of
non-Hodgkin’s lymphoma. She received intra-
venous ceftazidime perioperatively but had not
been treated with cytotoxic chemotherapy. She
had a peripheral blood leucocytosis of 22 000
cells/µl with 6% band forms. Flexible sig-
moidoscopy and biopsy demonstrated pseu-
domembranous colitis of the rectum and
sigmoid colon. Treatment was begun with both
intravenous metronidazole (500 mg, six
hourly) and oral vancomycin (250 mg, six
hourly). After five days she continued to suVer
from profuse diarrhoea and had a persistent
leucocytosis of 21 000 cells/µl. The patient’s
abdomen became distended and diVusely ten-
der. A plain abdominal radiograph showed an
ileus pattern with both small intestinal and
colonic dilatation. A computed tomogram
showed dilatation of the colon and the presence
of ascites (fig 1). Intravenous immunoglobulin
was administered (300 mg/kg). The diarrhoea
improved rapidly. After 36 hours her abdomi-
nal pain and distension had resolved and her
white blood cell count was normal at
9800 cells/µl. Treatment with metronidazole
and vancomycin was continued for a further 10
days. One month later she suVered a recur-
rence of diarrhoea and had a positive stool
cytotoxin assay. On this occasion she re-
sponded to treatment with oral metronidazole.
PATIENT 2
A 64 year old man underwent left upper lobec-
tomy for large cell lung cancer. Intravenous
vancomycin and ceftazidime were adminis-
tered postoperatively for the treatment of
pneumonia. Six days after surgery he devel-
oped diarrhoea, cramping abdominal pain, a
fever of 102°F (38.9°C), diVuse abdominal
tenderness, and abdominal distension. A stool
test for CdiYcile cytotoxin was positive and he
was treated with oral metronidazole (500 mg,
six hourly). An abdominal radiograph showed
thickening of the wall of the colon with thumb-
printing (fig 2A). Flexible sigmoidoscopy was
performed three days later because of increas-
ing abdominal pain and distension, and
showed pseudomembranous colitis (fig 2B).
Oral vancomycin (250 mg, six hourly) was ini-
tiated. Nine days later he showed no improve-
ment and had continuing diarrhoea, abdominal
discomfort, and intermittent fevers. Intra-
venous immunoglobulin (200 mg/kg) was ad-
ministered. Within 24 hours his diarrhoea and
fever resolved and did not recur.
C DIFFICILE ANTITOXIN ACTIVITY IN HUMAN
IMMUNOGLOBULIN PREPARATIONS
The rapid clinical response of these two
patients to intravenous administration of nor-
mal pooled human serum immunoglobulin led
us to test a variety of human IgG preparations
for neutralising antibodies against CdiYcile
toxins A and B.
All nine of the human immunoglobulin
preparations tested contained IgG against C
diYcile culture filtrate (fig 3A). Antibody levels
varied slightly with an approximately fourfold
diVerence in antibody titre between the prepa-
rations with the highest and lowest antibody
levels. We also measured IgG levels against
purified CdiYcile toxin A and toxin B. A
representative result for an immunoglobulin
preparation with mid-range anti-CdiYcile IgG
Figure 1: Computed tomogram of the abdomen of patient
1 showing dilated loops of colon (A) and the presence of
ascitic fluid (B).
Intravenous IgG for C diYcile colitis 367

levels is presented in fig 3B (this particular
preparation is identified by an arrow in fig 3A).
All preparations contained IgG against both C
diYcile toxin A and toxin B as measured by
ELISA.
Finally, we determined whether pooled
human immunoglobulin was capable of neu-
tralising the cytotoxic eVects of CdiYcile
toxins. All nine preparations neutralised C dif-
ficile culture filtrate cytotoxicity at IgG concen-
trations of 0.4–1.6 mg/ml. Control serum from
a healthy volunteer who lacked specific anti-
bodies against CdiYcile toxin A or toxin B
failed to neutralise the cytotoxicity of CdiYcile
culture filtrate in this assay.
Discussion
Most patients who develop CdiYcile diarrhoea
respond promptly to either oral metronidazole
or vancomycin.
126
Diarrhoea may recur when
these agents are discontinued but even then
almost always resolves quickly when antimicro-
bial therapy is resumed. Persisting diarrhoea
despite appropriate treatment with metronida-
zole and vancomycin, as occurred in both
patients in this report, is unusual. Both patients
also had severe colitis as evidenced by pseu-
domembrane formation, thickening of the
colonic wall, abdominal tenderness, and ab-
dominal distension. Severe, unresponsive pseu-
domembranous colitis may result in colonic
perforation, septicaemia, and death.
1 27–30
Colectomy may be life saving in these circum-
stances. However, many patients are consid-
ered unfit for colectomy because of advanced
age and severe coexisting medical problems.
Even those who are considered fit to undergo
colectomy for severe pseudomembranous coli-
tis have a mortality rate of approximately
50%.
27–30
Thus it was felt that intravenous
immunoglobulin treatment for unresponsive
pseudomembranous colitis was justified for the
two patients presented in this report. In both
instances there was rapid clinical improvement
immediately following immunoglobulin ad-
ministration.
Both patients in this report had recognised
risk factors for CdiYcile colitis including anti-
Figure 2: Plain abdominal radiograph of patient 2
showing colonic dilatation and thickening of the colonic
wall consistent with notable mucosal oedema (A).
Photomicrograph of sigmoidoscopic biopsy specimen
showing acutely inflamed colonic mucosa and an overlying
pseudomembrane (B).
Figure 3: C diYcile antitoxin activity in human
immunoglobulin preparations. (A) Anti-C diYcile culture
filtrate IgG levels measured by ELISA in commercially
available pooled normal human immunoglobulin
preparations. Three lots from each of three suppliers were
studied (Alpha Therapeutic Corporation (open triangle),
Armour Pharmaceutical Company (open square), and
Baxter Healthcare Corporation (open circle). (B) IgG
levels against C diYcile culture filtrate (open circle), toxin
A (open triangle), and toxin B (open square) in a
representative immunoglobulin preparation (denoted by the
arrow in A).
0.40
0.15
0.1
Total IgG (µg/ml)
Anti-C difficile IgG
(Optical density units)
0.25
0.8
0.40.2
0.20
0.30
0.35
1.6
A
0.35
0.05
0.1
Total IgG (µg/ml)
Anti-C difficile IgG
(Optical density units)
0.25
0.80.40.2
0.20
0.30
1.6
B
0.15
0.10
Culture filtrate
Toxin B
Toxin A
368 Salcedo, Keates, Pothoulakis, Warny, Castagliuolo, LaMont, et al

biotic treatment, admission to hospital, ad-
vanced age, malignancy, and recent major
surgery.
12273132
The elderly and those who
experience major surgery or trauma are known
to have impaired antibody responses against a
range of antigens and this may include reduced
CdiYcile antitoxin production.
14 16 33 34
Neither
patient had received cytotoxic chemotherapy at
the time of onset of diarrhoea but the first
patient had a lymphoma which may also be
associated with diminished antibody produc-
tion in response to antigenic challenge.
A number of previous studies have reported
low serum antibody levels against CdiYcile
toxins in patients with severe and prolonged C
diYcile colitis.
14 18–23
Elderly individuals, who
are most likely to develop CdiYcile diarrhoea,
may also have lower levels of neutralising anti-
toxin in their serum.
16
If inadequate antibody
production does indeed contribute to more
severe or more prolonged disease, intravenous
administration of preformed antitoxin may be
beneficial. Passive immunotherapy, either oral
or parenteral, is eVective in preventing C
diYcile enterocolitis in animals but little
information is available for humans.
14 24 35–37
The largest study to date is our report of
intravenous immunoglobulin administration to
children with recurrent CdiYcile
diarrhoea.
21 38 39
In that study we demonstrated
a notable increase in serum antitoxin A levels
following immunoglobulin administration. In
this study both patients were treated urgently
with immunoglobulin and, unfortunately,
serum was not saved to allow measurement of
their serum antitoxin antibody levels.
As CdiYcile colitis is toxin mediated we
assume that the immunoglobulin acts by bind-
ing and neutralising CdiYcile toxins. The
intravenous administration of 150 mg/kg of
immunoglobulin results in serum IgG levels of
greater than 5 mg/ml whereas toxin neutralis-
ing activity is evident in vitro at IgG concentra-
tions of approximately 1 mg/ml.
40
Thus neu-
tralising levels of IgG against CdiYcile toxins
are readily achieved in the blood following
immunoglobulin infusion. However, the pre-
cise mechanism of action of intravenous
immunoglobulin is still not entirely clear since,
to be eVective, IgG antitoxin must leave the
circulation and bind to toxins A and B within
the colonic lamina propria or intestinal
lumen.
91441
This may occur as a result of the
exudation of serum proteins across an inflamed
colonic mucosa.
Our first patient had a recurrence of CdiY-
cile diarrhoea four weeks after immunoglobulin
infusion. This may simply be the result of dis-
continuing metronidazole and vancomycin
three weeks earlier. However, the timing of her
relapse is also consistent with the expected
time course of degradation of intravenously
administered human IgG.
40
The animal and human studies discussed
earlier indicate that normal pooled human
immunoglobulin is a rational therapy for severe
CdiYcile colitis particularly in cases where
standard antimicrobial agents prove ineffective.
Immunoglobulin treatment carries some risks
but these are relatively small in the context of
acutely ill patients with severe colitis who are
often elderly and debilitated and already
receiving multiple medications. A controlled
prospective study would be needed to examine
properly the eYcacy of immunoglobulin in
these circumstances. However, the uncommon
and urgent nature of this condition makes such
a study logistically diYcult. The two cases
reported here suggest that passive immuno-
therapy with pooled normal human immu-
noglobulin may be a useful addition to metro-
nidazole and vancomycin and may hasten
recovery or avert colectomy in patients with
severe, refractory pseudomembranous colitis.
1 Kelly CP, Pothoulakis C, LaMont JT. Clostridium diYcile
colitis. N Engl J Med 1994; 330: 257–62.
2 McFarland LV, Mulligan ME, Kwok RY, Stamm WE.
Nosocomial acquisition of Clostridium diYcile infection.
N Engl J Med 1989; 320: 204–10.
3 Bongaerts GP, Lyerly DM. Role of toxins A and B in the
pathogenesis of Clostridium diYcile disease. Microb Pathog
1994; 17: 1–12.
4 Lyerly DM, Krivan HC, Wilkins TD. Clostridium diYcile:
its disease and toxins. Clin Microbiol Rev 1988; 1: 1–18.
5 Lyerly DM, Lockwood DE, Richardson SH, Wilkins TD.
Biological activities of toxins A and B of Clostridium diY-
cile. Infect Immun 1982; 35: 1147–50.
6 Riegler M, Sedivy R, Pothoulakis C, Hamilton G, Zacherl J,
Bischof G, et al. Clostridium diYcile toxin B is more potent
than toxin A in damaging human colonic epithelium in
vitro. J Clin Invest 1995; 95: 2004–11.
7 Triadafilopoulos G, Pothoulakis C, O’Brien MJ, LaMont
JT. DiVerential eVects of Clostridium diYcile toxins A and
B on rabbit ileum. Gastroenterology 1987; 93: 273–9.
8 Dove CH, Wang SZ, Price SB, Phelps CJ, Lyerly DM,
Wilkins TD, et al. Molecular characterization of the
Clostridium diYcile toxin A gene. Infect Immun 1990; 58:
480–8.
9 Kelly CP, Becker S, Linevsky JK, Joshi MA, O’Keane JC,
Dickey BF, et al. Neutrophil recruitment in Clostridium
diYcile toxin A enteritis in the rabbit. J Clin Invest 1994;
93: 1257–65.
10 Triadafilopoulos G, Pothoulakis C, Weiss R, Giampaolo C,
LaMont JT. Comparative study of Clostridium diYcile
toxin A and cholera toxin in rabbit ileum. Gastroenterology
1989; 97: 1186–92.
11 von Eichel-Streiber C, Laufenberg-Feldmann R, Sartingen
S, Schulze J, Sauerborn M. Cloning of Clostridium difficile
toxin B gene and demonstration of high N-terminal
homology between toxin A and B. Med Microbiol Immunol
(Berl) 1990; 179: 271–9.
12 Just I, Selzer J, Wilm M, von Eichel-Streiber C, Mann M,
Aktories K. Glucosylation of Rho proteins by Clostridium
diYcile toxin B. Nature 1995; 375: 500–3.
13 Just I, Wilm M, Selzer J, Rex G, von Eichel-Streiber C,
Mann M, et al. The enterotoxin from Clostridium diYcile
(ToxA) monoglucosylates the Rho proteins. J Biol Chem
1995; 270: 13932–6.
14 Kelly CP. Immune response to Clostridium diYcile. Eur J
Gastroenterol Hepatol 1996; 8: 1048–53.
15 Kelly CP, Pothoulakis C, Orellana J, LaMont JT. Human
colonic aspirates containing immunoglobulin A antibody to
Clostridium diYcile toxin A inhibit toxin A-receptor bind-
ing. Gastroenterology 1992; 102: 35–40.
16 Nakamura S, Mikawa M, Nakashio S, Takabatake M,
Okado I, Yamakawa K, et al. Isolation of Clostridium diY-
cile from the feces and the antibody in sera of young and
elderly adults. Microbiol Immunol 1981; 25: 345–51.
17 Viscidi R, Laughon BE, Yolken R, Bo-Linn P, Moench T,
Ryder RW, et al. Serum antibody response to toxins A and
B of Clostridium diYcile. J Infect Dis 1983; 148: 93–100.
18 Aronsson B, Granstrom M, Mollby R, Nord CE. Serum
antibody response to Clostridium diYcile toxins in patients
with Clostridium diYcile diarrhoea. Infection 1985; 13:
97–101.
19 Aronsson B, Granstrom M, Mollby R, Nord CE. Enzyme-
linked immunosorbent assay (ELISA) for antibodies to
Clostridium diYcile toxins in patients with pseudomem-
branous colitis and antibiotic-associated diarrhoea. J
Immunol Methods 1983; 60: 341–50.
20 Bacon AE III, Fekety R. Immunoglobulin G directed
against toxins A and B of Clostridium diYcile in the
general population and patients with antibiotic-associated
diarrhoea. Diagn Microbiol Infect Dis 1994; 18: 205–9.
21 Leung DY, Kelly CP, Boguniewicz M, Pothoulakis C, LaM-
ont JT, Flores A. Treatment with intravenously adminis-
tered gammaglobulin of chronic relapsing colitis induced
by Clostridium diYcile toxin. J Pediatr 1991; 118: 633–7.
22 Mulligan ME, Miller SD, McFarland LV, Fung HC, Kwok
RY. Elevated levels of serum immunoglobulins in asympto-
matic carriers of Clostridium diYcile. Clin Infect Dis 1993;
16 (suppl 4): S239–44.
Intravenous IgG for C diYcile colitis 369

23 Warny M, Vaerman JP, Avesani V, Delmee M. Human anti-
body response to Clostridium diYcile toxin A in relation to
clinical course of infection. Infect Immun 1994; 62: 384–9.
24 Kelly CP, Pothoulakis C, Vavva F, Castagliuolo I, Bostwick
EF, O’Keane JC, et al. Anti-Clostridium diYcile bovine
immunoglobulin concentrate inhibits cytotoxicity and
enterotoxicity of C. diYcile toxins. Antimicrob Agents
Chemother 1996; 40: 373–9.
25 Pothoulakis C, Barone LM, Ely R, Faris B, Clark ME,
Franzblau C, et al. Purification and properties of Clostrid-
ium diYcile cytotoxin B. J Biol Chem 1986; 261: 1316–21.
26 Kelly CP, LaMont JT. Treatment of Clostridium diYcile
diarrhoea and colitis. In: Wolfe MM, ed. Gastrointestinal
pharmacotherapy. Philadelphia: WB Saunders, 1993; 199–
212.
27 Medich DS, Lee KK, Simmons RL, Grubbs PE, Yang HC,
Showalter DP. Laparotomy for fulminant pseudomembra-
nous colitis. Arch Surg 1992; 127: 847–52.
28 Morris LL, Villalba MR, Glover JL. Management of
pseudomembranous colitis. Am Surg 1994; 60: 548–51.
29 Rubin MS, Bodenstein LE, Kent KC. Severe Clostridium
diYcile colitis. Dis Colon Rectum 1995; 38: 350–4.
30 Trudel JL, Deschenes M, Mayrand S, Barkun AN. Toxic
megacolon complicating pseudomembranous enterocolitis.
Dis Colon Rectum 1995; 38: 1033–8.
31 Bender BS, Bennett R, Laughon BE, Greenough WB III,
Gaydos C, Sears SD, et al. Is Clostridium diYcile endemic
in chronic-care facilities? Lancet 1986; ii: 11–3.
32 McFarland LV. Epidemiology of infectious and iatrogenic
nosocomial diarrhoea in a cohort of general medicine
patients. Am J Infect Control 1995; 23: 295–305.
33 Schwab R, Walters CA, Weksler ME. Host defense mecha-
nisms and aging. Semin Oncol 1989; 16: 20–7.
34 Weksler ME. Immune senescence. Ann Neurol 1994; 35
(suppl): S35–7.
35 Corthier G, Muller MC, Wilkins TD, Lyerly D, L’Haridon
R. Protection against experimental pseudomembranous
colitis in gnotobiotic mice by use of monoclonal antibodies
against Clostridium diYcile toxin A. Infect Immun 1991;
59: 1192–5.
36 Kim PH, Iaconis JP, Rolfe RD. Immunization of adult ham-
sters against Clostridium diYcile-associated ileocecitis and
transfer of protection to infant hamsters. Infect Immun
1987; 55: 2984–92.
37 Lyerly DM, Bostwick EF, Binion SB, Wilkins TD. Passive
immunization of hamsters against disease caused by
Clostridium diYcile by use of bovine immunoglobulin G
concentrate. Infect Immun 1991; 59: 2215–8.
38 Tjellstrom B, Stenhammar L, Eriksson S, Magnusson KE.
Oral immunoglobulin A supplement in treatment of
Clostridium diYcile enteritis [letter]. Lancet 1993; 341:
701–2.
39 Hassett J, Meyers S, McFarland L, Mulligan ME. Recurrent
Clostridium diYcile infection in a patient with selective
IgG1 deficiency treated with intravenous immune globulin
and Saccharomyces boulardii. Clin Infect Dis 1995; 20
(suppl 2): S266–8.
40 Pirofsky B, Campbell SM, Montanaro A. Individual patient
variations in the kinetics of intravenous immune globulin
administration. J Clin Immunol 1982; 2: 7S-14S.
41 Pothoulakis C, LaMont JT, Eglow R, Gao N, Rubins JB,
Theoharides TC, et al. Characterization of rabbit ileal
receptors for Clostridium diYcile toxin A. Evidence for a
receptor-coupled G protein. J Clin Invest 1991; 88: 119–25.
370 Salcedo, Keates, Pothoulakis, Warny, Castagliuolo, LaMont, et al