Treatment with intravenously administered gamma globulin of chronic relapsing colitis induced by Clostridium difficile toxin D o n a l d Y. M, L e u n g , MD, PhD. C i a r a n P, Kelly, MB, M a r k B o g u n i e w i c z , MD, C h a r a l a b o s Pothoulakis, MD, J, T h o m a s L a M o n t , MD, a n d A l e j a n d r o Flores, MD From the Division of Pediatric Allergy-immunology, National Jewish Center for Immunology and Respiratory Medicine, Denver, Colorado, and the Department of Pediatrics, Harvard Medical School, the Division of Immunology, The Children's Hospital, and the Section of Gastroenterology, Evans Memorial Department of Clinical Research, University Hospital, Boston University School of Medicine, Boston, Massachusetts

We tested the hypothesis that children with chronic relapsing colitis induced by Clostridium difficile toxin have defective antibody responses to C. difficile toxins as a cause of their underlying illness. Six such children were tested for serum IgG and IgA antibody to C. difficile toxin A. These six children had lower IgG anti-toxin A levels than 24 healthy children (p = 0.026) and 18 healthy adults (p = 0.0008). Five patients treated with 400 mg intravenously administered ~'-globulin per kilogram every 3 weeks had significant increases in IgG (p = 0.0t) but not IgA anti-toxin A (p = 0.406) levels, and all five had clinical resolution of their gastrointestinal symptoms as well as clearing of C. difficile cytotoxin B from their stools. These observations suggest that a dificiency of IgG anti-toxin A may predispose children to the development of chronic relapsing C. difficileinduced colitis. In such cases, intravenous ~-globulin therapy may be effective in producing clinical remission. (J PEDIATR199t;t18:633-7)

Clostridium difficile-induced antibiotic-associated colitis is toxin mediated.~ Pathogenic strains of C. difficite produce two protein exotoxins: toxin A, an enterotoxin, and toxin B, a cytotoxin. 1' 2 Toxin A induces fluid secretion into ligated rabbit ileal loops (enterotoxin effect) 24 and elicits an acute inflammatory response with granulocyte infiltration, epithelial cell necrosis, ulceration, and hemorrhagic edema in the intestinal mucosa. Purified toxin B does not cause fluid secretion or inflammatory damage in the rabbit or hamster

Supported in part by U.S. Public Health Service grant No. HL37260 and National Institutes of Health grant No. DK 34583. Submitted for publication Sept. 21, 1990; accepted Nov. 2, 1990. Reprint requests: Donald Y. M. Leung, MD, PhD, Department of Pediatrics, National Jewish Center for Immunology and Respiratory Medicine, 1400 Jackson St., Denver, CO 80206.

9/25/26496

intestine.3, 4 Thus toxin A appears to be the principal mediator of intestinal disease in colitis caused by C. difficile. Chronic C. difficile infection with relapse of colitis despite antibiotic therapy with metronidazole or vancomycin is a significant clinical problem in both adults and ELISA IVGG OD PBS-T

Enzyme-linked immunosorbent assay Intravenously administered ~/-globulin Optical density Phosphate-buffered saline solution with 0.05% Tween 20

children.5, 6 Several therapeutic maneuvers have been suggested, with inconsistent results. These efforts include alternative antibiotic regimens] the use of anion exchange resins to bind C. difficile toxin,8 and attempts to reestablish colonic flora by using Lactobacillus or fecal enemas.6' 9 We describe six patients with chronic relapsing colitis caused by

633

634

Leung et al.

The Journal of Pediatrics April 1991

Table I. Clinical and immune characteristics of patients with recurrent C. difficile-associated colitis Patient No.

Age (me)

Sex

Duration of diarrhea (me)

No. of positive stools for C. difficile cytotoxin B

Serum IgG (mg/dl)

1 37 M 32 2 812 2 36 M 17 6 363* 3 18 M 12 5 516 4 18 F 7 10 563 5 10 F 7 3 654 6 6 F 4 5 179" *Low irnmunoglobulinvalue for age or low tetanus toxoidtiter after boosterimmunizationor both.

C. difficile who had low serum C. difficile toxin A antibody levels, and their response to the intravenous administration of 3,-globulin. Data are presented to support the hypothesis that a low IgG antitoxin antibody titer may predispose children to the development of this condition. METHODS Study populatlon. Six children (median age 18 months, range 6 to 37 months; Table I) seen at Children's Hospital in Boston with chronic, relapsing C difficile-induced colitis were entered into the study after informed consent was obtained from parents. These six patients had a median duration of diarrhea of 7 months (range 4 to 32 months). The definition of chronic, relapsing colitis caused by C. difficile was as follows: (1) a history of broad-spectrum antibiotic use before the onset of diarrhea; (2) positive results on multiple stool C. difficile cytotoxin assays (Table I), in which the presence of C. difficile cytotoxin B in the stools was assayed by cytotoxic effect on cultured human fibroblasts 1; (3) the exclusion of other causes for diarrhea by stool cultures, colonoscopy with colonic biopsy, and, in most instances, upper gastrointestinal tract endoscopy and biopsy of the small intestine; and (4) relaPSe of colitis after courses of treatment with metronidazole, vancomycin, cholestyramine, or a combination of these drugs. In each case the children showed temporary symptomatic response while receiving antibiotic therapy, but a relapse occurred shortly after the antibiotic was discontinued. Two control groups were observed: 24 healthy children (median age 25 months, range 3 to 47 months) and 18 healthy adults (median age 57 years, range 23 to 91 years). These pediatric and adult control subjects had no history of antibiotic-associated diarrhea or colitis. Immune studies. Serum IgG, IgM, and IgA levels were measured by radial immunodiffusion with plates prepared by Kallestad Laboratories, Inc., Austin, Tex. ~~Serum IgG antibody titers to tetanus toxoid were measured by an enzyme-linked immunosorbent assay as previously described.ll

Serum IgM (mg/dl)

Serum IgA (mg/dl)

Tetanus toxoid titer (U/ml)

26* 61 65 130 69 24*

214 30 41 40 50 4*

0.6 1.8 0.5 1.3 I-0O Z LU r .J U I-a. O

tion for a reference positive serum run in parallel were 10.7% (IgG ELISA) and 10.8% (lgA ELISA). Statistical methods. The Mann-Whitney rank sum test was used for two group comparisons.

0 RESULTS Total lgG, IgM, and IgA levels and antibody titers, after immunization with the protein antigen tetanus toxoid, were assessed in six patients with chronic relapsing C. difficileinduced colitis. The results shown in Table I indicate that two of the six patients had low IgG antibody levels. Two had low IgM levels and one had low IgA levels. Furthermore, two patients did not mount a normal antibody response to tetanus toxoid (Table I). All patients had normal T-cell function as indicated either by a normal 48-hour delayed skin test reaction to tetanus toxoid or Candida organisms or by a normal in vitro proliferative response of peripheral blood mononuclear cells to mitogens and normal distribution of CD3-, CD4-, and CD8-bearing T cells (data not shown). These initial data suggested that a defect in humoral immunity could account for susceptibility to chronic relapsing C. difficile-induced colitis. We therefore determined IgG antibody levels to C. difficile-derived toxin A in sera from our six patients with colitis caused by C. difficile, 24 healthy children, and 18 healthy adults. Patients with chronic C. difficile-induced colitis had significantly lower levels of IgG anti-toxin A (median value 0.13 ELISA OD units) than those in normal children (median value 0.23 ELISA OD units; p = 0.03) or adults (median value 0.40 ELISA OD units; p = 0.0008) (Table If). In agreement with a previous study, 13 we found significantly higher IgG anti-toxin A levels in the adult control group than in pediatric control subjects (p = 0.01). We treated five of the patients with 400 mg IVGG per

I PRE- IVGG

I POST-

IVGG

Figure. Serum IgG antibody levels to C. difficile toxin A are shown for each of five patients before and after 1VGG therapy. Post-IVGG antibody levels shown are those measured just before third dose of 1VGG. Therapy with 1VGG resulted in a rise in antitoxin A antibody levels in all patients (p = 0.01).

kilogram every 3 weeks. No antibiotics were used during the period of IVGG administration. Patient 6 refused IVGG therapy and was lost to follow-up. The IVGG therapy was well tolerated without any observed side effects. All five patients had negative stool assays for C. difficile cytotoxin B (median time required for toxin clearance was two IVGG treatments) and a significant reduction in the frequency of bowel movements (median of seven bowel movements before IVGG and two bowel movements after the second IVGG dose; p < 0.01). Indeed, all clinical signs of colitis resolved during maintenance IVGG therapy. After IVGG therapy was stopped, one patient (No. 3) had a relapse of C. difficile-induced colitis, as indicated by a recurrence of diarrhea and positive stool assays for C. difficile cytotoxin. His symptoms were partially ameliorated with metronidazole and vancomycin therapy but recurred when therapy with these antibiotics was discontinued. Furthermore, this patient's IgG anti-toxin A levels dropped from 0.97 ELISA units with IVGG therapy to 0.20 ELISA units without therapy. Thus he was treated again with IVGG and again had resolution of his gastrointestinal symptoms and clearing of the C. difficile cytotoxin B within two doses of IVGG. The other four patients have had no symptoms with 1VGG

636

Leung et al.

therapy for between 4 and 6 months and are in the process of having their therapy stopped. Post-IVGG anti-toxin A IgG levels were obtained just before the third infusion of IVGG. Significantly higher levels of IgG anti-toxin A levels were obtained after infusion of IVGG (p -- 0.01) (Figure). In contrast, there was no significant change in IgA antitoxin levels during therapy (Table II). These results suggest that the clinical remission during IVGG therapy was caused by passive infusion of IgG antitoxin rather than by the development of active immunity against C. difficile toxin A. The IVGG preparations used in this study were found to have high levels of IgG anti-toxin A activity (1.93 ELISA OD units at 1 mg IgG per milliliter) but no detectable levels of IgA anti-toxin A activity. DISCUSSION

C. difficile toxin-induced colitis usually resolves spontaneously or after therapy with metronidazole or vancomycin.1, 6 Occasionally, however, patients have a chronic relapsing form of colitis caused by C. difficile that can be severely debilitating and unrelenting. Therapy with metronidazole or vancomycin provides only temporary relief in such cases; diarrhea and colitis recur when therapy with these antibiotics is discontinued. 1,6,7 We found that six children with chronic relapsing C. difficile-associated colitis had significantly lower mean antibody levels to C. difficile~lerived toxin A than those in normal children or adults. The potential relevance of the low anti-toxin A levels to the clinical illness is supported by the observation that several of these children had poor antigen-specific antibody responses after immunization with tetanus toxoid antigen. In a previous study, serum antibody to C. difficile toxin A was present in 64% of individuals older than 2 years of age but in only 19% of those younger than 2 years of age. 13 We also found significantly higher anti-toxin A antibody levels in adults than in children. These observations are compatible with the development of an antibody response to C. difficile toxin A, or a cross-reacting antigen, in the majority of persons during early childhood. In comparison with adult control subjects, many of our pediatric control subjects and all the children with chronic, relapsing C. difficile-associated colitis had low serum anti-toxin A levels. The distribution of antibody titers in the control pediatric group was bimodal; group 1 had low antibody levels and group 2 had antibody levels comparable to those in the adult group. As discussed above, low anti-toxin A antibody levels were expected in a proportion of pediatric control subjects because the prevalence of anti-toxin A antibody increases with age. The finding of tow anti-toxin A levels in patients with chronic relapsing C. difficile-assoeiated colitis was not expected, however, because colitis caused by C. difficile is

The Journal of Pediatrics April 1991 usually accompanied by a rise in serum antibody titer to toxin A. 13 This observation suggests that relapse in these children may be related to their failure to mount an appropriate antibody response on exposure to C. difficile toxin A. Although two of the six children in our study had low total IgG levels, chronic colitis caused by C. difficile has not been described as a complication of hypogammaglobulinemia. It is more likely, therefore, that in these children C. difficile toxin-induced colitis developed as a complication of the use of antibiotics required for treatment of upper respiratory tract infections. With the exception of patient 2, who had more than 20 ear infections and seven episodes of central line-related sepsis, none of the other patients had an unusual number of concomitant infections. Our rationale for IVGG therapy in these patients was based in part on previous reports that treatment with antitoxin antibody prevents fatal C. difficile-induced colitis in experimental animals. 14, 15 Because this study was open, the possibility exists that the C. difficile infection in these patients cleared spontaneously. This possibility appears unlikely for several reasons. First, these patients had had chronic relapsing C. difficile toxininduced colitis for up to 32 months, and yet all had resolution of symptoms with IVGG therapy. Second, despite a significant rise in IgG anti-toxin A levels after passive infusion of IVGG, we did not document any change in IgA anti-toxin A titers after this therapy. These data suggest that clinical improvement in symptoms was not accompanied by the attainment of active immunity against C. difficile. Finally, the serum of one patient who had a relapse of colitis caused by C. difficile when IVGG therapy was stopped was found again to contain low IgG anti-toxin A levels, and this patient responded to retreatment with IVGG. The mechanism of IVGG action in colitis caused by C. difficile is unknown. The pooled human IVGG contained high levels of anti-toxin A; therefore the therapeutic effect of IVGG in this disorder may have been related to the neutralization of toxin A by direct antitoxin activity. The clearance of cytotoxin B with IVGG therapy, however, suggests a broader effect possibly related to the presence of other antibodies to C. difficile in the IVGG preparation. Recent studies have also demonstrated that IVGG treatment can have antiinflammatory and immunomodulatory effects.16, 17These effects are usually seen when higher doses of IVGG are administered than the doses used in our current study. However, we cannot exclude the possibility that reduction of inflammation in the colon by IVGG could facilitate clearance of C. difficile toxins and recovery from illness. Our data suggest that IVGG may be an effective therapy for chronic relapsing C. difficile-induced colitis. Further controlled trials will be needed to confirm our observation.

Volume 118 Number 4, Part 1

IVGG f o r colitis caused by Clostridium difficile

REFERENCES 1. Lyerly DM, Krivan HC, Wilkias TD. Clostridium difficile: its disease and toxins. Clin Microbiol Rev 1988;1:1-18. 2. Triadafilopoulos G, Pothoulakis C, O'Brien M, LaMont JT. Differential effects of Clostridium difficile toxins A and B on rabbit ileum. Gastroenterology 1987;93:273-9. 3. Lyerly DM, Lockwood DE, Richardson SH, Wilkins TD. Biological activities of toxins A and B of Clostridium difficile. Infect Immun 1982;35:1147-50. 4. Mitchell T J, Ketley JM, Haslam SC, et al. Effect of toxins A and B on Clostridium difficile on rabbit ileum and colon. Gut 1986;27:78-85. 5. Sutphen JL, Grand R J, Flores A, Chang TW, Bartlett JG. Chronic diarrhea associated with Clostridium difficile in children. Am J Dis Child 1983;137:275-8. 6. Bartlett JG. Treatment of Clostridium difficile colitis. Gastroenterology 1985;89:1192-5. 7. Buggy BP, Fekety R, Silva J Jr. Therapy of relapsing Clostridium difficile associated diarrhea and colitis with the combination of vancomycin and rifampicin. J Clin Gastroenterol 1987;9:155-9. 8. Holmes R, Byrne WJ. Relapsing pseudomembranous colitis. J Pediatr Gastroenterol Nutr 1986;5:314-5. 9. Schwan A, Sjolin S, Trottestam 15, et al. Relapsing Clostridium difficile enterocolitis cured by rectal infusion of homologous feces. Lancet 1983;2:845-8.

10. Stiehm ER, Fudenberg HH. Serum levels of immune globulins in health and disease: a survey. Pediatrics 1966;37:715-27. 11. Leung DYM, Ambrosino DM, Arbeit RD, Newton JL, Geha RS. Impaired antibody responses in the hyperimmunoglobulinemia E syndrome. J Allergy Clin Immunol 1988 ;81:1082-7. 12. Sullivan NM, Pellter S, Wilkins TD. Purification and characterization of toxins A and B of Clostridium difficile. Infection 1982;35:1032-40. 13. Viscidi R, Loughon BE, Yolken R, et al. Serum antibody response to toxins A and B of Clostridium difficile. J Infect Dis 1983;148:93-100. 14. Allo M, Silva J, Fekety R, Rifkin GD, Waskin H. Prevention of clindamycin-induced colitis in hamsters by clostridium sordellii antitoxin. Gastroenterology 1979;76:351-5, 15. Kim PH, laconis JP, Rolfe RD. Immunization of adult hamsters against Clostridium difficile-associated ileocecitis and transfer of protection to infant hamsters. Infect Immun 1987;55:2984-92. 16. Leung DYM, Cotran RS, Kurt-Jones EZ, Burns JC, Newburger JW, Pober JS. Endothelial cell activation and increased interleukin 1 secretion in the pathogenesis of acute Kawasaki disease. Lancet 1989;2:1298-302. 17. Gelfand EW. Intervention in autoimmune disorders: creation of a niche for intravenous gammaglobulin therapy. Clin Immunol Immunopathol 1989;53:S1-7.

BOUND VOLUMES AVAILABLE TO SUBSCRIBERS Bound volumes of the 1991 issues of THE JOURNAL OF PEDIATRICS are available to subscribers (only) from the Publisher, at a cost of $52.00 for domestic, $71.64 for Canadian, and $68.00 for international subscribers, for Vol. 118 ( J a n u a r y - J u n e ) a n d Vol. 119 (July-December), shipping charges included. E a c h bound volume contains subject a n d a u t h o r indexes, a n d all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. T h e binding is durable b u c k r a m , with the Journal name, volume n u m b e r , and year stamped in gold on the spine. Payment must accompany all orders. C o n t a c t M o s b y - Y e a r Book, Inc., Subscription Services, 11830 Westline Industrial Dr., St. Louis, M O 63146-3318, U S A / 8 0 0 - 3 2 5 - 4 1 7 7 , ext. 4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular

Journal subscription.

637

Treatment with intravenously administered gamma globulin of chronic relapsing colitis induced by Clostridium difficile toxin.

We tested the hypothesis that children with chronic relapsing colitis induced by Clostridium difficile toxin have defective antibody responses to C. d...
431KB Sizes 0 Downloads 0 Views