An increasing number of allogeneic bone marrow transplantations (BMT) are being performed, in both developed and developing countries, for the treatment of patients with disorders of the hemopoietic system. Apart from regimen-related toxicities, these patients are at risk for infection with unusual and opportunistic pathogens because of immunosuppression and for developing immunological complications such as graft-versus-host disease (GVHD), all of which can affect the gastrointestinal tract (1–4).
There are very few reports of a systematic evaluation of the etiology of infectious diarrhea in patients undergoing BMT. Case reports and studies of small series of BMTs have reported enteritis caused by rotavirus (5,6), adenovirus (7), coxsackievirus (8), Acinetobacter (9), Cryptosporidium (10) and Giardia lamblia (11). Prospective studies of larger series have shown variable rates of intestinal infections in postBMT patients. A 40% incidence was reported in a study of 78 patients carried out in 1980–1981, with viruses and Clostridium difficile as the commonest cause of infections (12). Another study of 143 patients undergoing BMTs during 1981–1988 identified 36 infectious agents including bacteria, viruses, parasites and fungi (13). Enteric viruses were isolated from 19 patients in another prospective study of 94 patients (14). A smaller study of 12 pediatric patients reported a 52% incidence of infectious diarrhea as a result of viruses, parasites, and C. difficile (15). The largest and most systematic study to date followed 296 patients and found intestinal infections only in 20 episodes of diarrhea, with viruses and C. difficile the most commonly identified organisms (16).
No studies on the etiology of diarrhea have been carried out in developing countries so far. The objectives of this study were to determine the etiology of diarrhea after BMT with particular emphasis on infections, in a tropical environment, where most individuals have had multiple enteric infections before their admission to the hospital and where asymptomatic carriage of enteric pathogens is common (17,18).
PATIENTS AND METHODS
All patients who underwent allogeneic BMT from matched, related donors between September 1995 and March 1998 were included in the study. The 65 patients enrolled in this study were diagnosed to have thalassemia major (n=36), chronic myeloid leukemia (n=11), acute myeloid leukemia (n=8), aplastic anemia (n=6), acute lymphoblastic leukemia (n=1), Wiskott-Aldrich syndrome (n=1), congenital pure red cell aplasia (n=1), and paroxysmal nocturnal hemoglobinuria (n=1). Among the thalassemic patients, there were 24 males and 12 females, with an average age of 7.4 years. The 29 nonthalassemia patients were 17 males and 12 females, with an average age of 23.2 years.
Evaluation during study period.
During conditioning therapy, and in the posttransplantation period, signs and symptoms, drug usage, transfusions given, and the results of laboratory tests were recorded. Evaluations for mucositis and GVHD were carried out by the attending physician and supported by appropriate laboratory investigations.
Sample collection and examination.
All patients scheduled to undergo or undergoing BMTs were kept under microbiological surveillance from admission into the transplant unit, during conditioning therapy, and after the transplantation until recovery from neutropenia. Throat swabs and stool were collected from all patients at least once a week during the surveillance period, and these were cultured for aerobic bacterial pathogens and fungi. All organisms isolated from throat and stool had their antibiotic susceptibility patterns determined. Additional stool samples were collected at the same time from these patients, before transplantation and once a week for the first 4 weeks. If a patient developed diarrhea (defined as the passage of three or more stools of any consistency per day), additional samples were analyzed until cessation of diarrhea. Other than the surveillance cultures, a total of 336 stool specimens were obtained from 65 patients. The specimens were received in the laboratory and processed within 2 hr for bacteria, viruses, and parasites as described earlier (18). Briefly, bacteria were identified by standard enteric pathogen culture techniques, adherence assays, and toxin testing. Toxin testing and adherence assays were done for at least five colonies of Escherichia coli by techniques that have been described earlier (19). We tested for C. difficile toxin (CDT) using a cytotoxin assay by neutralization of cytotoxic effect by C. sordellii antiserum in HEp-2 cells (20). Viruses were identified by electron microscopy, latex agglutination for rotavirus (Meritec, Meridian Diagnostics, Cincinnati, OH), and culture in HT-29 and SkCo cells (21). Parasites were identified by examination of saline and iodine preparations and by modified acid-fast, Field’s, and modified trichrome staining, directly and after concentration.
Upper gastrointestinal endoscopy and biopsy were performed in three patients. Rectal biopsy specimens were taken for histology in 15 patients.
The chi-square test was used to assess significance of etiological association and differences in mortality between infected and uninfected patients.
Clinical course and outcome.
Four patients had diarrhea during the period of conditioning therapy. No pathogens were isolated from any of these patients. A total of 31 patients developed diarrhea in the posttransplantation period: 8 developed diarrhea in the first 20 days after transplantation, whereas 23 had diarrhea more than 20 days after transplantation. Enteric pathogens were identified in 17 patients with diarrhea. Pathogen isolation (1 of 8, 12.5%) in patients with diarrhea was low in the phase of neutropenia and higher 20 days after transplantation (16 of 23, 69.5%).
Mucositis was recorded in 14 patients, usually beginning only after the third day after transplantation and resolving within a few days, not later than 3 weeks after transplantation (Table 2). Twenty-six patients developed GVHD. Skin biopsy specimens were taken in 16 patients and showed GVHD in 12 patients. Rectal biopsy specimens from 15 patients showed mild GVHD in 5 patients, grade III GVHD in 2 patients, and grade IV GVHD in 3 patients. All three patients with grade IV GVHD on rectal biopsy died; CDT was identified from the stool filtrate of one of these patients. In the five rectal biopsy specimens in which there was no GVHD, acute proctitis was reported in two and acute colitis in two. One rectal biopsy specimen that showed acute proctitis was taken from a 5-year-old child from whom Shigella flexneri had been isolated before transplantation, who continued to have intermittent diarrhea up to day 73 after transplantation and subsequently died. CDT was identified in the stool of one patient with acute colitis, but this patient was treated with vancomycin and recovered. Eight patients had GVHD affecting more than one organ system.
During the course of the study, 21 transplant patients died, 3 in the first 20 days after transplantation, 11 between 20 and 100 days, and 7 at more than 100 days after transplantation. Eight had thalassemia, six had acute myeloid leukemia, four had aplastic anemia, and three had chronic myeloid leukemia. Twelve (38.7%) of 31 patients who had diarrhea after transplantation died. In four patients, the terminal events could be attributed directly to gut-related disease, GVHD, or infection.
Pretransplantation screening of stool showed the presence of enteric pathogens in 19 patients (Table 1). The organisms isolated were mainly parasites (n=10) and viruses (n=6). None of the 19 patients had diarrhea in the pretransplantation period, but 8 developed diarrhea in the posttransplantation period. At the time of diarrhea, 4 of the 8 had the same pathogens as had been identified before transplantation. No organisms were seen in the remaining four patients.
In 11 patients (7 with parasites, 3 with caliciviruses, 1 with adenovirus, 2 with enteroaggregative E. coli [EAggEC], and 1 with Candida), although putative enteric pathogens were identified before transplantation, the patients did not develop diarrhea at any time. The patient with adenovirus identified by electron microscopy on pretransplantation screening did not have diarrhea but developed hematuria 40 days after transplantation, and adenovirus was observed in the urine.
Posttransplantation examination of stool revealed enteric pathogens in 31 patients (Table 1). The organisms isolated were bacteria (n=20), viruses (n=11), and parasites (n=6).
One patient had Cryptosporidium identified both before and after transplantation but had diarrhea only in the posttransplantation period. Two other patients excreted Cryptosporidium oocysts in the posttransplantation period, and both had intermittent diarrhea during the third and fourth weeks, which subsequently subsided. The other patients in whom parasites were identified did not have diarrhea. Microsporidia were not identified in any patient. There was no significant association between parasitic infections, diarrhea morbidity, development of GVHD, or overall mortality in this group of patients.
Of 10 patients with viral infections after transplantation, 5 of 6 rotavirus infections were symptomatic. Both patients with adenovirus had diarrhea associated with pain and bleeding more than a month after transplantation: one patient recovered, while the other had gut and liver GVHD, developed septicemia and pulmonary aspergillosis, and died. One patient with diarrhea during the fourth week after transplantation was diagnosed with cytomegalovirus (CMV) enteritis histologically but also developed GVHD and disseminated aspergillosis and died.
Acute infection as a result of CMV occurred in 10 patients, in whom either CMV IgM or polymerase chain reaction was positive, or CMV inclusion bodies were seen on histopathology with associated active infection. There was no significant association between enteric viral infection and development of GVHD or overall mortality, but CMV disease, not limited to the gastrointestinal tract, was associated with increased mortality (P <0.05).
Symptomatic bacterial infections were caused by S. flexneri, enterotoxigenic E. coli (ETEC), enteropathogenic E. coli (EPEC), enterohemorrhagic E. coli (EHEC), and C. difficile. A total of nine patients had CDT identified in stool in the posttransplantation period, of whom one did not have diarrhea. One patient who had CDT identified from stool developed gastrointestinal bleeding and died 5 weeks after transplantation. Three patients had CDT identified from stool at the time of diarrhea 3–4 weeks after transplantation, were treated with vancomycin, recovered, and then developed aspergillosis (disseminated=1, pulmonary=2) and died. Two patients with CDT in stool and diarrhea in the fourth posttransplantation week were treated with vancomycin and recovered. There was a significant association between posttransplantation bacterial infection and morbidity caused by diarrhea (P <0.02) and overall mortality (P <0.01), but not with development of GVHD.
Pretransplantation infections were identified in 19 patients but did not result in diarrhea. Unlike posttransplantation infections, the majority of the infections were viral and parasitic, with pathogens that are known to be carried asymptomatically. According to the protocols followed in the transplant unit, when a putative enteric pathogen is identified on pretransplantation screening, appropriate therapy is instituted promptly. However, the transplantation is not delayed because 5 days of appropriate antimicrobial therapy is completed before the neutropenic phase.
In this study, 48% of patients undergoing BMT developed diarrhea. Mucositis seemed to be the commonest cause of diarrhea during the first 20 days after transplantation (Table 2). Early GVHD was seen in four patients, and only two diarrheal patients had demonstrable pathogens in the first 20 days after transplantation. In two patients with GVHD, immunosuppressive therapy resulted in a temporary response, followed by enteric superinfection with C. difficile and shiga-like toxin-producing EHEC, with a fatal outcome.
Most patients with parasitic infections, other than four patients with Cryptosporidium (one of whom had Isospora as well), did not develop diarrhea. Giardia lamblia is known to cause severe diarrhea coincident with marrow ablation (11), but in the single patient with asymptomatic giardiasis in this study, the cysts were identified 4 weeks after transplantation, after the period of posttransplantation epithelial recovery.
Among the viral agents of gastroenteritis, rotavirus caused significant clinical disease in children up to 12 years in this setting, with only one asymptomatic infection. Other studies have reported an incidence of 10% for rotaviral infections after BMT (12,22), similar to the 12.3% incidence reported here. Infections with caliciviruses were asymptomatic, therefore further identification of these viruses was not attempted. In studies from the U.S., symptomatic infections caused by coxsackieviruses have been reported as a major cause of morbidity and mortality in BMT patients, especially in outbreaks of infection (8), but none were seen here. Infection with enteric adenoviruses has been reported in 5–10% of patients undergoing BMT (7,14). In our study, the incidence was 4.6% (3/65), and one patient did not have diarrhea but had hematuria. Although CMV enteritis was diagnosed in only one patient, there was evidence of acute or reactivated CMV infection in 10 patients. CMV infection of the gastrointestinal tract can be most difficult to document, because it involves subepithelial tissues rather than epithelial cells and may involve intestinal segments unreachable by endoscopy (23). In an earlier, postmortem study on 10 bone marrow transplant recipients from this institution, CMV infection was seen in 3 patients, affecting the esophageal mucosa and lungs (24), indicating that CMV is likely to be a major cause of morbidity in this setting.
A total of 25 potentially enteropathogenic bacteria were isolated from 21 patients, of whom 17 were symptomatic (Table 1). All except one patient with C. difficile had diarrhea, which usually responded to vancomycin. Antibiotic therapy was not useful in other bacterial infections. In developed countries, enteric pathogens such as Salmonella, Shigella, and Campylobacter are rarely found in the protected environment of transplant units (12,25,26), but unlike tropical settings (17,18) these are also rarely identified in the normal human population. In this study, both Salmonella and Shigella were isolated from posttransplantation patients. Although both patients with Shigella died after transplantation, both had been treated with ciprofloxacin, to which the isolates were sensitive and there was no evidence that gastrointestinal infection caused the mortality. However, nosocomial enteric infections caused by these and other pathogens have been reported previously in pediatric patients in this setting (27), although these patients, unlike those undergoing BMT, did not have their food sterilized before consumption.
Previous studies have not examined the role of HEp-2 cell adherent E. coli, ETEC, and EHEC in posttransplantation diarrhea. In this study, only one isolate produced the localized pattern, and this did not belong to the classical enteropathogenic groups. Three isolates produced the aggregative pattern, but EAggEC infection was not associated with diarrhea in two of three patients.
A prospective study on patients undergoing BMT (16) reported that GVHD was a more common (48%) cause of diarrhea than enteric infection (13%). This is not similar to our findings, where after day 20, infections (15 of 25, 60%) were seen in association with diarrhea more frequently than GVHD (11 of 25, 44%). However, given the complexity of physiological changes in these patients, it is difficult to assign causality to a single factor in isolation.
The high prevalence of bacterial enteric pathogens may be a reflection of the high level of asymptomatic colonization with enteric pathogens that is present in the community (17,18), where 20–50% of asymptomatic subjects can harbor potentially pathogenic enteric organisms. Identification of colonization in an asymptomatic individual requires repeated sample collection, because the numbers of organisms excreted may be low or infrequently found in the stool, depending on the life cycle (28). In this study, the collection of multiple samples in all patients may be one factor responsible for the high rate of identification of putative pathogens in asymptomatic patients. Although Indian patients undergoing BMT are likely to be at increased risk for development of gastrointestinal infections, they may be more resistant to infection by organisms to which they have been previously exposed multiple times.
In a previous study, in which enteric pathogens were isolated from 40% of patients undergoing BMT, it was found that patients with enteric infections had a four times higher risk of dying than patients who did not have infections (12). This is also borne out by this study, in which 18 of 39 patients with enteric pathogens died and there were only 3 deaths in 26 patients without enteric infections. Although, enteric infection seemed to be the primary cause of death in only two patients, it is possible that patients with infections of the gastrointestinal tract may be more susceptible to infection of other systems as well. More detailed studies of infections of other organ systems may help in elucidating the role of different types of infections in morbidity and mortality in patients undergoing BMT.
In recent years, studies have examined the relationship of gastrointestinal and oropharyngeal microflora and the incidence of GVHD, with some investigators finding that decontamination of these regions can lead to a significant reduction in the incidence of infections and GVHD (29,30). Some of the protocols available are promising and should be tested in our population for prevention of morbidity caused by gastrointestinal and other disease.
1. Tutschka PJ. Infections and immunodeficiency in bone marrow transplantation. Pediatr Infect Dis J 1988; 7 (suppl 5): S22.
2. McDonald GB, Shulman HM, Sullivan KM, Spencer GD. Intestinal and hepatic complications of human bone marrow transplantation. Part I. Gastroenterology 1986; 90: 460.
3. Epstein RJ, McDonald GB, Sale GE, Shulman HM, Thomas ED. The diagnostic accuracy of the rectal biopsy in acute graft-versus-host disease: a prospective study of thirteen patients. Gastroenterol 1980; 78: 764.
4. Slavin RE, Santos GW. The graft versus host reaction in man after bone marrow transplantation: pathology, pathogenesis, clinical features and implication. Clin Immunol Immunopathol 1973; 1: 472.
5. Willoughby RE, Wee SB, Yolken RH. Non-group: a rotavirus infection associated with severe gastroenteritis in a bone marrow transplant patient. Pediatr Infect Dis J 1988; 7: 133.
6. Kruger W, Stockschlader M, Zander AR. Transmission of rotavirus diarrhea in a bone marrow transplantation unit by a hospital worker. Bone Marrow Transplant 1991; 8: 507.
7. Shields AF, Hackman RC, Fife KH, Corey L, Meyers JD. Adenovirus infections in patients undergoing bone-marrow transplantation. N Engl J Med 1985; 312: 529.
8. Townsend TR, Bolyard EA, Yolken RH, et al. Outbreak of Coxsackie A1 gastroenteritis: a complication of bone marrow transplantation. Lancet 1982; i: 820.
9. Nagler A, Pavel L, Naparstek E, Muggia-Sullam M, Slavin S. Typhlitis occurring in autologous bone marrow transplantation. Bone Marrow Transplant 1992; 9: 63.
10. Gentile G, Venditti M, Micozzi A, et al. Cryptosporidiosis in patients with hematologic malignancies. Rev Infect Dis 1991; 13 (5): 842.
11. Bromiker R, Korman SH, Or R, et al. Severe giardiasis in two patients undergoing bone marrow transplantation. Bone Marrow Transplant 1989; 4: 701.
12. Yolken RH, Bishop CA, Townsend TR, et al. Infectious gastroenteritis in bone-marrow transplant recipients. N Engl J Med 1982; 306: 1009.
13. Aplin MS, Weiner R, Graham Pole J, Hiemenz J, Elfenbein G, Oblon DJ. Enterocolitis in allogeneic marrow transplant patients [Abstract]. Exp Hematol 1990; 18: 696.
14. Troussard X, Bauduer F, Gallet E, et al. Virus recovery from stools of patients undergoing bone marrow transplantation. Bone Marrow Transplant 1993 12: 573.
15. Blakely S, Barnes GL, Bishop RF, Ekert H. Infectious diarrhea in children undergoing bone marrow transplantation. Aust N Z J Med 1989; 19: 31.
16. Cox GJ, Matsui SM, Lo RS, et al. Etiology and outcome of diarrhea after marrow transplantation: a prospective study. Gastroenterology 1994; 107: 1398.
17. Rajan DP, Mathan VI. The prevalence of enteric bacterial pathogens in a healthy population in southern India. J Med Microbiol 1986; 22: 93.
18. Mathew M, Mathan MM, Mani K, et al. The relationship of microbial pathogens to acute infectious diarrhoea of childhood. J Trop Med Hyg 1991; 94: 253.
19. Kang G, Mathan MM, Mathan VI. Evaluation of a simplified HEp-2 cell adherence assay for Escherichia coli
isolated from south Indian children with acute diarrhea and controls. J Clin Microbiol 1995; 33: 2204.
20. Allen SD, Baron EJ. Clostridium. In: Balows A, Hausler WJ Jr, Herrmann KL, Isenberg HD, Shadomy HJ, eds. Manual of clinical microbiology, 5th Edition. Washington, DC: American Society for Microbiology, 1995: 505.
21. Winston DJ, Gale RP, Meyer DV, Young LS, the UCLA Bone Marrow Transplantation Group. Infectious complications of human bone marrow transplantation. Medicine 1979; 58: 1.
22. Patel JR, Daniel J, Mathan VI. A comparison of the susceptibility of three human gut tumor-derived differentiated epithelial cell lines, primary monkey kidney cells and human rhabdomyosarcoma cell line to 66-prototype strains of human enteroviruses. J Virol Methods 1985 12: 209.
23. Winston DJ, Ho WG, Champlin RE. Cytomegalovirus infections after allogeneic bone marrow transplantation. Rev Infect Dis 1990; 12 (suppl 7): S776.
24. Ramani P, Krishnaswami H, Srivastava A, Dennison D, Chandy M. Infections in bone marrow transplant recipients in South India: a post-mortem study. J Trop Med Hyg 1994; 97: 354.
25. McDonald GB, Shulman HM, Sullivan KM, Spencer GD. Intestinal and hepatic complications of human bone marrow transplantation. Part II. Gastroenterology 1986; 90: 770.
26. Meyers JD, Thomas ED. Infection complicating bone marrow transplantation. In: Rubin RH, Young LS, eds. Clinical approach to infections in the compromised host. New York: Plenum, 1981: 507.
27. Kang G, Mathew MS, Rajan DP, et al. Prevalence of intestinal parasites in rural South Indians. Trop Med Int Health 1998; 3: 70.
28. Kamalaratnam CN, Kang G, Kirubakaran C, et al. A prospective study of nosocomial enteric pathogen acquisition in hospitalized children in South India. J Trop Pediatr 2001; 47: 46.
29. Vossen JM, Heidt PJ. Gnotobiotic measures for the prevention of acute graft-versus-host disease. In: Burakoff SJ, Deeg HJ, eds. Graft-versus-host disease: immunology, pathophysiology and treatment. New York: Marcel Dekker, 1990: 403.
30. Liu KL, Herbecht R, Lioure B, et al. Effects of bacterial naso-oropharyngeal colonisation on acute graft-versus-host disease in patients undergoing bone marrow transplantation. Bone Marrow Transplant 1995; 15: 65.