- Email: [email protected]
A prospective study of viral and mycoplasma infections in chronic inflammatory bowel disease
GASTROENTEROLOGY 1990;98:549-553
ALIMENTARY
TRACT
A Prospective Study of Viral and Mycoplasma Infections in Chronic Inflammatory Bowel Disease H. 0. KANGRO, S. K. F. CHONG, and J. A. WALKER-SMITH
A. HARDIMAN,
Departments of Virology and Child Health, St. Bartholomew’s London, United Kingdom
Seventy-two children with chronic inflammatory bowel disease were investigated for infections with various viruses and Mycoplasma pneumoniae, Chlamydia psittaci, and Coxiella burnetii to determine whether these pathogens are associated with acute onset exacerbations. Altogether 54 infections were identified serologically, of which 23 (42.6%) were associated with exacerbations. This corresponded to 24.2% of the recorded exacerbations during the study period. The respiratory pathogens accounted for 59.3% of the infections and 43.6% of these were associated with gastrointestinal symptoms. This is consistent with the observation that up to 40% of the exacerbations were associated with symptoms of antecedent or concurrent infection, most commonly involving the respiratory tract. Rubella virus, Epstein-Barr virus, and adenovirus were associated with acute exacerbations in 5 children. Thus, common pathogens were frequently associated with exacerbations and account for a large proportion of the commonly reported symptoms of a concurrent infection. The possible causal relationship between these pathogens and exacerbation of inflammatory bowel disease is discussed. Reactivation of latent herpesviruses was identified in 4 children with active disease and indicates that the converse relationship may also occur.
cute episodes of exacerbation of chronic inflammatory bowel disease (CIBD) in children are frequently preceded or accompanied by symptoms of infection, most commonly of the respiratory tract (RT). Mee and Jewel1 (1) found that RT symptoms are the most common antecedent event in patients with exacerbations, occurring in 60% of cases. Rarely, however, is the infecting agent identified and therefore it is difficult to assess whether any significant association
A
R. B. HEATH,
Hospital, West Smithfield,
exists between specific infections and exacerbation of CIBD. It has been reported that bacterial and viral gastroenteritis can precipitate exacerbation of CIBD (2,3) but this does not account for the more common presentation of RT symptoms together with symptoms of exacerbation. Viruses are the most common cause of RT infections in children and many different viruses can cause RT symptoms. This prospective study was, therefore, undertaken to determine how frequently identifiable virus infections occurred in association with acute episodes of exacerbation or relapses of CIBD in children, and if such episodes are related to particular infectious agents. Materials
and Methods
Patients and Serum Specimens Between October 1981 and )une 1983, 85 children with CIBD (64 with Crohn’s disease, 18 with ulcerative colitis, and 3 with unclassified CIBD) aged 0.5-16.3 yr were recruited into the study. Seventy-two of these children were followed up prospectively for periods up to 20 mo [total 80.4 man-years to determine whether any association exists between intercurrent infections and acute exacerbations. The children routinely attend the Paediatric Inflammatory Bowel Disease Clinic at St. Bartholomew’s Hospital every 3-4 mo irrespective of disease activity and also as required on clinical grounds, i.e., exacerbation of symptoms. At each visit recent symptoms of infection and episodes of exacerbation or increased disease activity were recorded. Blood samples were collected at each visit for routine investigations as well as viral serology. Additional blood specimens
Abbreviations used in this paper: CIBD, chronic inflammatory bowel disease: RSV, respiratory syncytial virus; RT, respiratory tract. o 1999 by the American Gastroenterological Association OOM-5oa5/90R3.oo
550 KANGRO ET AL.
were collected whenever a child was admitted to the hospital for assessment or surgery. A total of 385 specimens were collected for serology during the study period. In addition, 97 blood specimens (mainly single or paired] were obtained for viral serology from 69 control children of similar age and sex distribution. These were children investigated in the Department of Child Health for conditions other than CIBD; most frequently celiac disease, food allergies, or gastroenteritis, and who were having routine blood tests.
GASTROENTEROLOGY
Vol. 98, No. 3
tuating antibody levels that varied from twofold to fourfold; therefore only antibody increases in consecutive sera of greater than eightfold were considered diagnostic. A diagnosed infection was regarded as symptomatic if the symptoms occurred within the time span of no more than 1 mo between paired sera showing an antibody increase or conversion, except in cases where measles, rubella, or chicken pox was diagnosed clinically. Results
Clinical Investigations The diagnosis of Crohn’s disease and ulcerative colitis was based on a system of histologic diagnostic criteria for classification of chronic inflammatory bowel disease in childhood, after initial assessment by clinical, radiologic, and endoscopic criteria (4,5). The evaluation of the children during the study was conducted by two of us on an outpatient basis. The method of assessing the response or lack of response of the disease to the treatment regime was quantified by a “disease activity index” in each patient at a particular time. This appraisal took into consideration the following factors: (a) number of stools per day, (b) abdominal pain rating, [c] state of general well-being, [d) loss of body weight, and [e] specific clinical signs of disease activity, i.e., fever, mouth ulcers, arthritis, uveitis, anal fissure, fistula, and abscess. An exacerbation of the disease is defined by a change in the symptomatology pattern at that particular visit with an increase in the disease activity index, and a resultant increase or change in the previous medication to control the disease. Oral prednisolone at doses of 1-2 mg/kg body wt per day
is restarted during a relapse of the disease. This is then gradually reduced to 0.25 mg/kg body wt per day after 2 mo with the aim of maintaining quiescent activity or clinical well-being.
Serological
Tests
Routine complement fixation tests were used to test serum specimens for antibodies to herpes simplex virus, measles, mumps, respiratory syncytial virus (RSV), parainfluenza virus, adenovirus, influenza A and B viruses, Mycoplasma pneumoniae, Chlamydia psittaci, and Coxiella burnetii. Radioimmunoassay (6) was used to test for antibodies against varicella-zoster virus, and single radial hemolysis (77, in conjunction with M-antibody capture radioimmunoassay, was used to test for rubella infection (8). In addition, sera from some children were tested for immunoglobulin G and immunoglobulin M class antibodies to Epstein-Barr virus capsid antigen by the indirect immunofluorescent antibody test at the Virology Department, St. George’s Hospital, London. A recent infection was diagnosed by the demonstration of virus-specific immunoglobulin M antibodies [rubella, Epstein-Barr virus], seroconversion, or a significant increase in the antibody level in consecutive specimens. During the follow-up period many children showed fluc-
Seasonal Incidence Study Population
of Exacerbations
in the
During the study period 158 episodes of acuteonset gastrointestinal (GI) symptoms were recorded in the 85 children with CIBD. No marked seasonal pattern was observed although an increased frequency of exacerbations was observed during the winter months, in April-May and in July-August. The winter and spring increases coincided with increased incidence of respiratory tract symptoms reported by the children [Figure 1). During the peak winter months 30%~40% of exacerbations were accompanied by RT symptoms. These peaks closely coincided with RSV and influenza B epidemics in the United Kingdom but not with other respiratory virus infections during the study period (data from the Communicable Diseases Surveillance Centre). identified
Infections
in the Study Population
A total of 54 infections were diagnosed in the children with CIBD (0.7 per man-year] and 23 (42.6%) of these were reported to be associated with GI symptoms (Table 1). Of the total 158 recorded exacerbations, 95 occurred in 56 children at a time when appropriately timed [within 1 mo) blood samples were obtained to enable a serological diagnosis of a concurrent infection. Thus 24.2% (23 of 95) of the exacerbations that were investigated serologically were associated with infections.
16
c___*
O.l.Mnpmrn
.........
“RTI
-
1
ls8l
lee2
“RTl Mllo.1.
Mnparr
lwa
Figure 1. Seasonal frequency of reported symptomatic RT infections, with or without GI symptoms, compared with the seasonal frequency of acute episodes of exacerbations in children with CIBD.
VIRUSES AND INFLAMMATORY
March 1999
Table 1. Summary of Infections Serologically Diagnosed in Children With Chronic Inflammatory Bowel Disease
Infectious Agents Respiratory Influenza A Influenza B RSV PIV M. pneumoniae Total Systemic Rubella Measles Mumps vsz Epstein-Barr virus Total
Infections [n)
Infections associated with acute GI symptoms lnl
a
4 (31
3
7 9 4 4 32
5 (31
3
4 (31
3
4 (21
2
2 I21 19
14
7b 4 0 3
15
Other HSV Adenovirus C. burnetti Chl. psittaci Total Overall total
Patients with symptoms of infection” (n)
3
2
3 3 -
0 -
0 -
3
0
0
1 10
1
1
3
2
3
4
0
2
2 -
-
54
Patients requiring change in medication lnl
5
6
34
23
11
HSV, herpes simplex virus; PIV, parainfluenza virus; RSV, respiratory syncytial virus; VZV, varicella-zoster virus. “Numbers in parentheses show combined respiratory tract infection and GI symptoms. %cluding three possible vaccinations.
Respiratory tract infections. Altogether 32 respiratory infections were diagnosed by serology (Table 1). Nineteen of these were reported as symptomatic and 14 (43.8%) were accompanied by acute GI symptoms. In all cases except one the GI symptoms occurred in association with symptoms of RT infection. The occurrence of GI symptoms was not specifically related to an infectious agent, although mycoplasma and parainfluenza virus infections were most commonly associated with GI symptoms. Infections with these two agents also necessitated a change in the medication more frequently than did infections with influenza virus or RSV (Table 1). Thus, of four mycoplasma infections three were associated with GI symptoms. Two of these occurred in children who were in remission and necessitated steroid therapy. The third infection occurred in a child who was on steroids and who 5 mo previously had a parainfluenza infection, also accompanied by an exacerbation. The GI symptoms associated with influenza viruses and RSV were generally mild and settled spontaneously. Only 3 children, of whom 2 were on a maintenance dosage of
BOWEL DISEASE
551
steroids, required intensified therapy. Overall, however, the incidence and severity of the GI symptoms associated with RT infection were not obviously related to the level of medication. Systemic infections. Of the 15 systemic infections that were diagnosed (Table l), 10 were symptomatic and three of these were associated with acute GI symptoms. Seven rubella infections were diagnosed but three of these were in girls aged 11-14 yr and may have been vaccine related. Two symptomatic rubella infections were accompanied by exacerbations. One of these occurred in a child treated with salazopyrin and colifoam and precipitated a severe exacerbation that required steroid therapy whereas the other, in a child who had previously been in remission, was mild and settled spontaneously. Three varicella-zoster virus infections (two chickenpox, one shingles) and four measles infections were diagnosed but none of these were associated with exacerbation of CIBD, in spite of the fact that five of these occurred in children who were on steroid therapy. One Epstein-Barr virus infection occurred in a child with previously well-controlled Crohn’s disease. He presented dramatically with lower respiratory tract symptoms, prolonged fever, a rash, and lymphadenopathy, which was accompanied by a severe exacerbation of the inflammatory bowel disease. Investigations showed neutropenia with atypical monocytosis, and positive Paul-Bunnel and Epstein-Barr virusspecific serology. The exacerbation of his CIBD initially settled on increased doses of steroids but within 1 mo further exacerbations followed that required further incremental doses of steroids for 6 mo. Other infections. Three adenovirus infections were identified (Table 1) in previously well children, of whom 2 coincidentally had sudden onset of bloody diarrhea. In both children the diarrhea settled rapidly, although within 2 mo 1 of the children showed increasing disease activity requiring steroid therapy. Four herpes simplex virus infections were diagnosed (Table l] of which three were identified in children with cold sores or other mouth ulceration. All 3 children had active disease and were on steroids. The oral lesions developed 7-14 days after the onset of exacerbation, unlike the symptoms of the respiratory and systemic infections that occurred concomitant with, or immediately preceding, the GI symptoms. All four infections occurred in children who were previously known to be seropositive and thus proven to be reactivated rather than primary herpes simplex virus infections. No infections with Chl. psittaci or C. burnetii were identified.
GASTROENTEROLOGY
552 KANGRO ET AL.
Multiple infections. Five children experienced more than one infection during the study period (Table 2). Exacerbations accompanied each of the infections in 3 children but none of the infections in the other 2 children, suggesting a predisposing factor in some children that is independent of the type of infecting agent. Infections
in Control Children
There was no significant difference in the prevalence of antibodies to the infectious agents between the control and the CIBD children, when first tested on entry into the study. Unfortunately, only a very limited follow-up was possible in these children, but in 15 cases serial specimens were obtained during the study period and altogether 8 infections were diagnosed (0.9 per man-year). Six of these were respiratory infections (2 influenza, 3 RSV, and 1 mycoplasma] and 2 were systemic (mumps and varicellazoster virus). Only 1 (12.5%) infection was associated with GI symptoms: an RSV infection that presented with a combination of RT symptoms and diarrhea in a child who was not otherwise being investigated for GI disease. Thus, infections appeared to be more commonly associated with GI symptoms in children with CIBD (42.6%) than in the controls (12.5%), although this is not statistically significant (p = 0.1). Discussion The results of this study confirm the frequent temporal association of acute exacerbations and infections in children with CIBD. In some months up to Table 2. Details of Five Children
With Chronic Inflammatory Bowel Disease Who Experienced Multiple Infections Medication
Child No. 1
2
3 4
5
at Time of Infection Nil Nil Nil Salazopyrin and colifoam Salazopyrin and colifoam Nil Nil Salazopyrin Salazopyrin Steroids Steroids
GI, gastrointestinal; syncytial virus.
Date
Infection
GI Symptoms
6/82 l/83 2/83
Rubella Mycoplasma Influenza A RSV
Nil Nil Nil Exacerbation
Rubella
Exacerbation
Influenza B RSV Rubella Influenza A PIV Mycoplasma
Exacerbation Exacerbation Nil Nil Exacerbation Exacerbation
12/81
6/82
2/82 la/82 5/82 12/82 11/82 4/83
PIV. parainfluenza
virus; RSV, respiratory
Vol. 98, No. 3
40% of the children reported symptoms of infection, most commonly of the RT, concurrently with exacerbations. By comparison, Mee and Jewel1 (1) found this association in 60% of adult patients with CIBD but the stricter criteria used in our study probably accounts for the difference. These criteria were applied to reduce the chance of recording purely coincidental associations, which are likely to occur especially during epidemic peak periods. Serological investigations identified infections in 24.2% of all recorded exacerbations. Infections with the respiratory pathogens were most commonly diagnosed. These pathogens are not generally thought to affect the GI tract, as infection does not usually involve spread beyond the RT. However, one of the RSV infections in the control children was associated with diarrhea and previous studies during an influenza outbreak in 1982 in normal school children showed that in addition to RT symptoms, 28% had diarrhea, 52% had nausea, and 58% had abdominal pain (9). Presumably such symptoms result from general toxic effects of the infections. Clinically these symptoms could mimic an exacerbation and therefore in some of the children in our study the GI symptoms may have been similarly related to the infections. This may explain why the GI symptoms in most RSV and influenza infections were mild and transient, lasting only a few days. Conversely, with four of the respiratory virus infections and the mycoplasma infections the symptoms were more severe, prolonged, and required medication. In these cases, therefore, the respiratory pathogens may have been involved in precipitating acute exacerbation of CIBD. Such an association possibly represents a more severe response in an already diseased gut to the toxic effects of infection. The response in patients with CIBD could thus range from no effect on the underlying disease to a relapse. The toxic effects are most likely mediated by soluble factors released during the immune response to the agent, or perhaps by immune complexes as proposed by Mee and Jewel1 (1). In systemic infections the opportunity exists for the virus itself to affect the GI tract as viremia is a feature of their pathogenesis. However, only rubella and Epstein-Barr virus were found to be associated with acute exacerbations. In 2 cases these were severe and followed within 3 days the onset of the symptoms of infection. It seems likely that the virus infections in these cases were directly involved as a precipitating cause. The involvement of an immunopathologic process is possible as previously has been proposed for the gut ulceration seen in some patients with cytomegalovirus infection (10). A different association between CIBD and virus infections was identified in 4 children who showed reactivation of herpes simplex virus and 1 child with zoster. Reactivation of herpes simplex
VIRUSES
March 1990
virus, unlike varicella-zoster virus, is not uncommon in childhood but in the 3 children who developed cold sores the symptoms appeared shortly after the onset of exacerbation and therefore presumably the increased disease activity, or the steroid therapy, was involved in reactivating the virus. We have also found that reactivation of cytomegalovirus frequently occurs in children with active CIBD (unpublished observations]. From our study, and others, it appears that viral and bacterial infections are often temporally associated with acute exacerbation of CIBD. Perhaps as suggested previously (2) most, if not all, episodes are triggered by extrinsic factors, particularly infections. In the children in our study most of the common virus infections were implicated and may have accounted for nearly 25% of exacerbations. We did not investigate the role of enteric viruses, although two adenovirus infections were documented that were associated with severe GI symptoms. Gebhard et al. (3) showed that in adult patients with CIBD up to 10% of acute exacerbations are associated with enteric virus infections and the proportion in children would not be expected to be less. The association between common infections and exacerbation of CIBD is intriguing because of the diverse nature of the agents and the varied responses in different patients. Although an infection may provide the initial trigger other factors appear to be involved. The involvement of a patient-related factor is supported by the findings in children with multiple infections, but no recognizable risk factor, such as medication or type of CIBD, could be identified. Clearly, further studies involving assessment by colonoscopy and histology are needed to better define the varied responses to infection seen in different patients. This may eventually provide information that could influence the management of patients.
AND INFLAMMATORY
BOWEL
DISEASE
553
References 1.
2. 3.
4.
5.
6.
7.
8.
9. 10.
Mee AS, Jewel1 DP. Factors inducing relapse in inflammatory bowel disease. Br Med J [Clin Res] 1978;2:801-2. Gorbach SL. Viral infections and inflammatory bowel disease. Gastroenterology 1982;83:1318-9. Gebhard RL, Greenberg HB, Singh N, et al. Acute viral enteritis and exacerbations of inflammatory bowel disease. Gastroenterology 1982;83:1207-9. Chong SKF, Bartram C, Campbell CA, Williams CB, Blackshaw AJ, Walker-Smith JA. Chronic inflammatory bowel disease in childhood. Br Med J [Clin Res] 1982;284:101-3. Chong SKF, Blackshaw AJ, Boyle S, Williams CB. Walker-Smith JA. Historical diagnosis of chronic inflammatory bowel disease in childhood. Gut 1985;26:55-9. Campbell-Benzie A, Kangro HO, Heath RB. The development and evaluation of a solid-phase radioimmunoassay procedure for the determination of susceptibility to varicella. J Virol Methods 1981;2:149-58. Kurtz JB, Mortimer PP. Mortimer PR, Morgan-Capner P, Shafi MS, White GBB. Rubella antibody measured by radial haemolysis characteristics and performance of a simple screening method for use in diagnostic laboratories. J Hyg (Lond) 1980;84: 213-22. Tedder RS. Yao JL, Anderson MJ. The production of monoclonal antibodies to rubella haemagglutinin and their use in antibody-capture assays for rubella-specific IgM. J Hyg (Lond) 1982;88:335-50. C.D.C. Influenza activity-Mississippi, United States, worldwide. Morbidity and Mortality Weekly Report 1984;33:131. Cooper HS, Raffeusperger EC, Jonas L, Fitts WT. Cytomegalovirus inclusions in patients with ulcerative colitis and toxic dilation requiring colonic resection. Gastroenterology 1977;72: 1253-6.
Received October 2.1986. Accepted April 19.1989. Address requests for reprints to: Dr. H. 0. Kangro, Virology Department, St. Bartholomew’s Hospital, Third Floor, 5X/53 Bartholomew Close, West Smithfield, London EClA 7BE, United Kingdom. This study was supported by grant G8121655 SB from the Medical Research Council.
ALIMENTARY
TRACT
A Prospective Study of Viral and Mycoplasma Infections in Chronic Inflammatory Bowel Disease H. 0. KANGRO, S. K. F. CHONG, and J. A. WALKER-SMITH
A. HARDIMAN,
Departments of Virology and Child Health, St. Bartholomew’s London, United Kingdom
Seventy-two children with chronic inflammatory bowel disease were investigated for infections with various viruses and Mycoplasma pneumoniae, Chlamydia psittaci, and Coxiella burnetii to determine whether these pathogens are associated with acute onset exacerbations. Altogether 54 infections were identified serologically, of which 23 (42.6%) were associated with exacerbations. This corresponded to 24.2% of the recorded exacerbations during the study period. The respiratory pathogens accounted for 59.3% of the infections and 43.6% of these were associated with gastrointestinal symptoms. This is consistent with the observation that up to 40% of the exacerbations were associated with symptoms of antecedent or concurrent infection, most commonly involving the respiratory tract. Rubella virus, Epstein-Barr virus, and adenovirus were associated with acute exacerbations in 5 children. Thus, common pathogens were frequently associated with exacerbations and account for a large proportion of the commonly reported symptoms of a concurrent infection. The possible causal relationship between these pathogens and exacerbation of inflammatory bowel disease is discussed. Reactivation of latent herpesviruses was identified in 4 children with active disease and indicates that the converse relationship may also occur.
cute episodes of exacerbation of chronic inflammatory bowel disease (CIBD) in children are frequently preceded or accompanied by symptoms of infection, most commonly of the respiratory tract (RT). Mee and Jewel1 (1) found that RT symptoms are the most common antecedent event in patients with exacerbations, occurring in 60% of cases. Rarely, however, is the infecting agent identified and therefore it is difficult to assess whether any significant association
A
R. B. HEATH,
Hospital, West Smithfield,
exists between specific infections and exacerbation of CIBD. It has been reported that bacterial and viral gastroenteritis can precipitate exacerbation of CIBD (2,3) but this does not account for the more common presentation of RT symptoms together with symptoms of exacerbation. Viruses are the most common cause of RT infections in children and many different viruses can cause RT symptoms. This prospective study was, therefore, undertaken to determine how frequently identifiable virus infections occurred in association with acute episodes of exacerbation or relapses of CIBD in children, and if such episodes are related to particular infectious agents. Materials
and Methods
Patients and Serum Specimens Between October 1981 and )une 1983, 85 children with CIBD (64 with Crohn’s disease, 18 with ulcerative colitis, and 3 with unclassified CIBD) aged 0.5-16.3 yr were recruited into the study. Seventy-two of these children were followed up prospectively for periods up to 20 mo [total 80.4 man-years to determine whether any association exists between intercurrent infections and acute exacerbations. The children routinely attend the Paediatric Inflammatory Bowel Disease Clinic at St. Bartholomew’s Hospital every 3-4 mo irrespective of disease activity and also as required on clinical grounds, i.e., exacerbation of symptoms. At each visit recent symptoms of infection and episodes of exacerbation or increased disease activity were recorded. Blood samples were collected at each visit for routine investigations as well as viral serology. Additional blood specimens
Abbreviations used in this paper: CIBD, chronic inflammatory bowel disease: RSV, respiratory syncytial virus; RT, respiratory tract. o 1999 by the American Gastroenterological Association OOM-5oa5/90R3.oo
550 KANGRO ET AL.
were collected whenever a child was admitted to the hospital for assessment or surgery. A total of 385 specimens were collected for serology during the study period. In addition, 97 blood specimens (mainly single or paired] were obtained for viral serology from 69 control children of similar age and sex distribution. These were children investigated in the Department of Child Health for conditions other than CIBD; most frequently celiac disease, food allergies, or gastroenteritis, and who were having routine blood tests.
GASTROENTEROLOGY
Vol. 98, No. 3
tuating antibody levels that varied from twofold to fourfold; therefore only antibody increases in consecutive sera of greater than eightfold were considered diagnostic. A diagnosed infection was regarded as symptomatic if the symptoms occurred within the time span of no more than 1 mo between paired sera showing an antibody increase or conversion, except in cases where measles, rubella, or chicken pox was diagnosed clinically. Results
Clinical Investigations The diagnosis of Crohn’s disease and ulcerative colitis was based on a system of histologic diagnostic criteria for classification of chronic inflammatory bowel disease in childhood, after initial assessment by clinical, radiologic, and endoscopic criteria (4,5). The evaluation of the children during the study was conducted by two of us on an outpatient basis. The method of assessing the response or lack of response of the disease to the treatment regime was quantified by a “disease activity index” in each patient at a particular time. This appraisal took into consideration the following factors: (a) number of stools per day, (b) abdominal pain rating, [c] state of general well-being, [d) loss of body weight, and [e] specific clinical signs of disease activity, i.e., fever, mouth ulcers, arthritis, uveitis, anal fissure, fistula, and abscess. An exacerbation of the disease is defined by a change in the symptomatology pattern at that particular visit with an increase in the disease activity index, and a resultant increase or change in the previous medication to control the disease. Oral prednisolone at doses of 1-2 mg/kg body wt per day
is restarted during a relapse of the disease. This is then gradually reduced to 0.25 mg/kg body wt per day after 2 mo with the aim of maintaining quiescent activity or clinical well-being.
Serological
Tests
Routine complement fixation tests were used to test serum specimens for antibodies to herpes simplex virus, measles, mumps, respiratory syncytial virus (RSV), parainfluenza virus, adenovirus, influenza A and B viruses, Mycoplasma pneumoniae, Chlamydia psittaci, and Coxiella burnetii. Radioimmunoassay (6) was used to test for antibodies against varicella-zoster virus, and single radial hemolysis (77, in conjunction with M-antibody capture radioimmunoassay, was used to test for rubella infection (8). In addition, sera from some children were tested for immunoglobulin G and immunoglobulin M class antibodies to Epstein-Barr virus capsid antigen by the indirect immunofluorescent antibody test at the Virology Department, St. George’s Hospital, London. A recent infection was diagnosed by the demonstration of virus-specific immunoglobulin M antibodies [rubella, Epstein-Barr virus], seroconversion, or a significant increase in the antibody level in consecutive specimens. During the follow-up period many children showed fluc-
Seasonal Incidence Study Population
of Exacerbations
in the
During the study period 158 episodes of acuteonset gastrointestinal (GI) symptoms were recorded in the 85 children with CIBD. No marked seasonal pattern was observed although an increased frequency of exacerbations was observed during the winter months, in April-May and in July-August. The winter and spring increases coincided with increased incidence of respiratory tract symptoms reported by the children [Figure 1). During the peak winter months 30%~40% of exacerbations were accompanied by RT symptoms. These peaks closely coincided with RSV and influenza B epidemics in the United Kingdom but not with other respiratory virus infections during the study period (data from the Communicable Diseases Surveillance Centre). identified
Infections
in the Study Population
A total of 54 infections were diagnosed in the children with CIBD (0.7 per man-year] and 23 (42.6%) of these were reported to be associated with GI symptoms (Table 1). Of the total 158 recorded exacerbations, 95 occurred in 56 children at a time when appropriately timed [within 1 mo) blood samples were obtained to enable a serological diagnosis of a concurrent infection. Thus 24.2% (23 of 95) of the exacerbations that were investigated serologically were associated with infections.
16
c___*
O.l.Mnpmrn
.........
“RTI
-
1
ls8l
lee2
“RTl Mllo.1.
Mnparr
lwa
Figure 1. Seasonal frequency of reported symptomatic RT infections, with or without GI symptoms, compared with the seasonal frequency of acute episodes of exacerbations in children with CIBD.
VIRUSES AND INFLAMMATORY
March 1999
Table 1. Summary of Infections Serologically Diagnosed in Children With Chronic Inflammatory Bowel Disease
Infectious Agents Respiratory Influenza A Influenza B RSV PIV M. pneumoniae Total Systemic Rubella Measles Mumps vsz Epstein-Barr virus Total
Infections [n)
Infections associated with acute GI symptoms lnl
a
4 (31
3
7 9 4 4 32
5 (31
3
4 (31
3
4 (21
2
2 I21 19
14
7b 4 0 3
15
Other HSV Adenovirus C. burnetti Chl. psittaci Total Overall total
Patients with symptoms of infection” (n)
3
2
3 3 -
0 -
0 -
3
0
0
1 10
1
1
3
2
3
4
0
2
2 -
-
54
Patients requiring change in medication lnl
5
6
34
23
11
HSV, herpes simplex virus; PIV, parainfluenza virus; RSV, respiratory syncytial virus; VZV, varicella-zoster virus. “Numbers in parentheses show combined respiratory tract infection and GI symptoms. %cluding three possible vaccinations.
Respiratory tract infections. Altogether 32 respiratory infections were diagnosed by serology (Table 1). Nineteen of these were reported as symptomatic and 14 (43.8%) were accompanied by acute GI symptoms. In all cases except one the GI symptoms occurred in association with symptoms of RT infection. The occurrence of GI symptoms was not specifically related to an infectious agent, although mycoplasma and parainfluenza virus infections were most commonly associated with GI symptoms. Infections with these two agents also necessitated a change in the medication more frequently than did infections with influenza virus or RSV (Table 1). Thus, of four mycoplasma infections three were associated with GI symptoms. Two of these occurred in children who were in remission and necessitated steroid therapy. The third infection occurred in a child who was on steroids and who 5 mo previously had a parainfluenza infection, also accompanied by an exacerbation. The GI symptoms associated with influenza viruses and RSV were generally mild and settled spontaneously. Only 3 children, of whom 2 were on a maintenance dosage of
BOWEL DISEASE
551
steroids, required intensified therapy. Overall, however, the incidence and severity of the GI symptoms associated with RT infection were not obviously related to the level of medication. Systemic infections. Of the 15 systemic infections that were diagnosed (Table l), 10 were symptomatic and three of these were associated with acute GI symptoms. Seven rubella infections were diagnosed but three of these were in girls aged 11-14 yr and may have been vaccine related. Two symptomatic rubella infections were accompanied by exacerbations. One of these occurred in a child treated with salazopyrin and colifoam and precipitated a severe exacerbation that required steroid therapy whereas the other, in a child who had previously been in remission, was mild and settled spontaneously. Three varicella-zoster virus infections (two chickenpox, one shingles) and four measles infections were diagnosed but none of these were associated with exacerbation of CIBD, in spite of the fact that five of these occurred in children who were on steroid therapy. One Epstein-Barr virus infection occurred in a child with previously well-controlled Crohn’s disease. He presented dramatically with lower respiratory tract symptoms, prolonged fever, a rash, and lymphadenopathy, which was accompanied by a severe exacerbation of the inflammatory bowel disease. Investigations showed neutropenia with atypical monocytosis, and positive Paul-Bunnel and Epstein-Barr virusspecific serology. The exacerbation of his CIBD initially settled on increased doses of steroids but within 1 mo further exacerbations followed that required further incremental doses of steroids for 6 mo. Other infections. Three adenovirus infections were identified (Table 1) in previously well children, of whom 2 coincidentally had sudden onset of bloody diarrhea. In both children the diarrhea settled rapidly, although within 2 mo 1 of the children showed increasing disease activity requiring steroid therapy. Four herpes simplex virus infections were diagnosed (Table l] of which three were identified in children with cold sores or other mouth ulceration. All 3 children had active disease and were on steroids. The oral lesions developed 7-14 days after the onset of exacerbation, unlike the symptoms of the respiratory and systemic infections that occurred concomitant with, or immediately preceding, the GI symptoms. All four infections occurred in children who were previously known to be seropositive and thus proven to be reactivated rather than primary herpes simplex virus infections. No infections with Chl. psittaci or C. burnetii were identified.
GASTROENTEROLOGY
552 KANGRO ET AL.
Multiple infections. Five children experienced more than one infection during the study period (Table 2). Exacerbations accompanied each of the infections in 3 children but none of the infections in the other 2 children, suggesting a predisposing factor in some children that is independent of the type of infecting agent. Infections
in Control Children
There was no significant difference in the prevalence of antibodies to the infectious agents between the control and the CIBD children, when first tested on entry into the study. Unfortunately, only a very limited follow-up was possible in these children, but in 15 cases serial specimens were obtained during the study period and altogether 8 infections were diagnosed (0.9 per man-year). Six of these were respiratory infections (2 influenza, 3 RSV, and 1 mycoplasma] and 2 were systemic (mumps and varicellazoster virus). Only 1 (12.5%) infection was associated with GI symptoms: an RSV infection that presented with a combination of RT symptoms and diarrhea in a child who was not otherwise being investigated for GI disease. Thus, infections appeared to be more commonly associated with GI symptoms in children with CIBD (42.6%) than in the controls (12.5%), although this is not statistically significant (p = 0.1). Discussion The results of this study confirm the frequent temporal association of acute exacerbations and infections in children with CIBD. In some months up to Table 2. Details of Five Children
With Chronic Inflammatory Bowel Disease Who Experienced Multiple Infections Medication
Child No. 1
2
3 4
5
at Time of Infection Nil Nil Nil Salazopyrin and colifoam Salazopyrin and colifoam Nil Nil Salazopyrin Salazopyrin Steroids Steroids
GI, gastrointestinal; syncytial virus.
Date
Infection
GI Symptoms
6/82 l/83 2/83
Rubella Mycoplasma Influenza A RSV
Nil Nil Nil Exacerbation
Rubella
Exacerbation
Influenza B RSV Rubella Influenza A PIV Mycoplasma
Exacerbation Exacerbation Nil Nil Exacerbation Exacerbation
12/81
6/82
2/82 la/82 5/82 12/82 11/82 4/83
PIV. parainfluenza
virus; RSV, respiratory
Vol. 98, No. 3
40% of the children reported symptoms of infection, most commonly of the RT, concurrently with exacerbations. By comparison, Mee and Jewel1 (1) found this association in 60% of adult patients with CIBD but the stricter criteria used in our study probably accounts for the difference. These criteria were applied to reduce the chance of recording purely coincidental associations, which are likely to occur especially during epidemic peak periods. Serological investigations identified infections in 24.2% of all recorded exacerbations. Infections with the respiratory pathogens were most commonly diagnosed. These pathogens are not generally thought to affect the GI tract, as infection does not usually involve spread beyond the RT. However, one of the RSV infections in the control children was associated with diarrhea and previous studies during an influenza outbreak in 1982 in normal school children showed that in addition to RT symptoms, 28% had diarrhea, 52% had nausea, and 58% had abdominal pain (9). Presumably such symptoms result from general toxic effects of the infections. Clinically these symptoms could mimic an exacerbation and therefore in some of the children in our study the GI symptoms may have been similarly related to the infections. This may explain why the GI symptoms in most RSV and influenza infections were mild and transient, lasting only a few days. Conversely, with four of the respiratory virus infections and the mycoplasma infections the symptoms were more severe, prolonged, and required medication. In these cases, therefore, the respiratory pathogens may have been involved in precipitating acute exacerbation of CIBD. Such an association possibly represents a more severe response in an already diseased gut to the toxic effects of infection. The response in patients with CIBD could thus range from no effect on the underlying disease to a relapse. The toxic effects are most likely mediated by soluble factors released during the immune response to the agent, or perhaps by immune complexes as proposed by Mee and Jewel1 (1). In systemic infections the opportunity exists for the virus itself to affect the GI tract as viremia is a feature of their pathogenesis. However, only rubella and Epstein-Barr virus were found to be associated with acute exacerbations. In 2 cases these were severe and followed within 3 days the onset of the symptoms of infection. It seems likely that the virus infections in these cases were directly involved as a precipitating cause. The involvement of an immunopathologic process is possible as previously has been proposed for the gut ulceration seen in some patients with cytomegalovirus infection (10). A different association between CIBD and virus infections was identified in 4 children who showed reactivation of herpes simplex virus and 1 child with zoster. Reactivation of herpes simplex
VIRUSES
March 1990
virus, unlike varicella-zoster virus, is not uncommon in childhood but in the 3 children who developed cold sores the symptoms appeared shortly after the onset of exacerbation and therefore presumably the increased disease activity, or the steroid therapy, was involved in reactivating the virus. We have also found that reactivation of cytomegalovirus frequently occurs in children with active CIBD (unpublished observations]. From our study, and others, it appears that viral and bacterial infections are often temporally associated with acute exacerbation of CIBD. Perhaps as suggested previously (2) most, if not all, episodes are triggered by extrinsic factors, particularly infections. In the children in our study most of the common virus infections were implicated and may have accounted for nearly 25% of exacerbations. We did not investigate the role of enteric viruses, although two adenovirus infections were documented that were associated with severe GI symptoms. Gebhard et al. (3) showed that in adult patients with CIBD up to 10% of acute exacerbations are associated with enteric virus infections and the proportion in children would not be expected to be less. The association between common infections and exacerbation of CIBD is intriguing because of the diverse nature of the agents and the varied responses in different patients. Although an infection may provide the initial trigger other factors appear to be involved. The involvement of a patient-related factor is supported by the findings in children with multiple infections, but no recognizable risk factor, such as medication or type of CIBD, could be identified. Clearly, further studies involving assessment by colonoscopy and histology are needed to better define the varied responses to infection seen in different patients. This may eventually provide information that could influence the management of patients.
AND INFLAMMATORY
BOWEL
DISEASE
553
References 1.
2. 3.
4.
5.
6.
7.
8.
9. 10.
Mee AS, Jewel1 DP. Factors inducing relapse in inflammatory bowel disease. Br Med J [Clin Res] 1978;2:801-2. Gorbach SL. Viral infections and inflammatory bowel disease. Gastroenterology 1982;83:1318-9. Gebhard RL, Greenberg HB, Singh N, et al. Acute viral enteritis and exacerbations of inflammatory bowel disease. Gastroenterology 1982;83:1207-9. Chong SKF, Bartram C, Campbell CA, Williams CB, Blackshaw AJ, Walker-Smith JA. Chronic inflammatory bowel disease in childhood. Br Med J [Clin Res] 1982;284:101-3. Chong SKF, Blackshaw AJ, Boyle S, Williams CB. Walker-Smith JA. Historical diagnosis of chronic inflammatory bowel disease in childhood. Gut 1985;26:55-9. Campbell-Benzie A, Kangro HO, Heath RB. The development and evaluation of a solid-phase radioimmunoassay procedure for the determination of susceptibility to varicella. J Virol Methods 1981;2:149-58. Kurtz JB, Mortimer PP. Mortimer PR, Morgan-Capner P, Shafi MS, White GBB. Rubella antibody measured by radial haemolysis characteristics and performance of a simple screening method for use in diagnostic laboratories. J Hyg (Lond) 1980;84: 213-22. Tedder RS. Yao JL, Anderson MJ. The production of monoclonal antibodies to rubella haemagglutinin and their use in antibody-capture assays for rubella-specific IgM. J Hyg (Lond) 1982;88:335-50. C.D.C. Influenza activity-Mississippi, United States, worldwide. Morbidity and Mortality Weekly Report 1984;33:131. Cooper HS, Raffeusperger EC, Jonas L, Fitts WT. Cytomegalovirus inclusions in patients with ulcerative colitis and toxic dilation requiring colonic resection. Gastroenterology 1977;72: 1253-6.
Received October 2.1986. Accepted April 19.1989. Address requests for reprints to: Dr. H. 0. Kangro, Virology Department, St. Bartholomew’s Hospital, Third Floor, 5X/53 Bartholomew Close, West Smithfield, London EClA 7BE, United Kingdom. This study was supported by grant G8121655 SB from the Medical Research Council.