CLINICAL CHARACTERISTICS AND ENDOSCOPIC FINDINGS ASSOCIATED WITH BLASTOCYSTIS HOMINIS IN HEALTHY ADULTS

INTRODUCTION

Since the discovery of Blastocystis hominis in the early 1900s, only its morphology has been extensively studied and well described.¹ Many of its aspects, including mode of transmission, pathogenicity, culture characteristics, taxonomy, life cycle, biochemistry and molecular biology, remain unclear.¹ Udkow and Markell suggested that an intestinal tract that is abnormal for any reason might provide conditions suitable for proliferation of B. hominis.² Furthermore, case reports and series have suggested a pathogenic role of B. hominis in causing intestinal inflammation.^3,4 However, there were also reports indicating otherwise.^5–7 Even in the symptomatic patients who harbored this protozoan, Zuckerman and others found that there was typically no evidence of significant intestinal inflammation and impaired barrier function of the intestinal mucosa.⁶ Despite great controversy, controlled studies of the association between intestinal lesions and B. hominis are lacking. This case-control study examined the association between acquisition of B. hominis and mucosal lesions revealed by upper gastrointestinal (UGI) endoscopy and sigmoidoscopy.

MATERIALS AND METHODS

Individuals who had received a routine physical examination at Taipei Veterans General hospital from November 2000 to October 2002 were identified from the computer database of medical records. The study was reviewed and approved by the Department of Medical Research and Education at Taipei Veterans General Hospital. Cases were defined as patients who were positive for B. hominis but negative for other parasites on fecal examination. Controls were selected from those without any parasite detected in the stool specimens, and were matched for age (within five years), sex, and date of examination with the positive cases. If there were more than two individuals who met the criteria for inclusion in the control group, two of them were randomly selected. The medical records of these positive cases and controls were retrospectively reviewed. Data including clinical characteristics, findings of UGI endoscopy, sigmoidoscopy, and laboratory tests were collected for analyses. Analyzed laboratory data included tumor markers and surface antigen of hepatitis B virus (HBsAg) positivity examined by radioimmunoassay, and antibody against hepatitis C virus (HCV) and human immunodeficiency virus (HIV) detected by enzyme-linked immunosorbent assay. The stool specimens were investigated by a direct wet film method for detection of leukocytes, ova, and parasites. The presence of B. hominis was reported after confirmation by two technicians using the merthiolate-iodine-formalin method.⁸ Guaiac testing was used to detect occult blood in the stool. Among those who had received gastric biopsy, the presence of Helicobacter pylori was suggested if this bacterium was revealed in the histologic examination or if a urease test result of a gastric specimen⁹ was positive. Eosinophilia was defined as ≥ 0.5 × 10⁹ eosinophils/L.¹⁰ A chi-square test with Yates’ correction or Fisher’s exact test were used to compare differences in discrete variables, and the Student’s t-test or the Mann-Whitney rank sum test was used to analyze continuous variables as appropriate. A multivariate analysis with logistic regression was performed to identify factors independently associated with B. hominis. All the analyses were performed with SPSS version 11 (SPSS Institute, Chicago, IL). A P value < 0.05 was considered statistically significant.

RESULTS

During the two-year study period, 101 (1.13%) of 8,941 individuals who underwent a routine physical examination were identified as harboring B. hominis. Two of the 101 individuals were excluded, one because of detection of Giardia lamblia and the other because of detection of nonpathogenic Ameba concomitantly with B. hominis in the fecal examination. Of the remaining 99 positive cases, 77 were male and 22 were female, with a mean ± SD age of 57.24 ± 10.26 years. A total of 193 matched controls were selected and their characteristics are shown in Table 1.

Nonspecific gastrointestinal (GI) symptoms presented in 12 cases during the examination included nausea, abdominal discomfort, flatulence, and diarrhea. However, a comparable percentage of controls also experienced these similar GI symptoms (P = 0.938). There were no significant differences in medical history, including diabetes mellitus (11 of 99 versus 17 of 193; P = 0.672), between the two groups. Study of the characteristics of the 28 individuals with diabetes revealed no significant difference in fasting glucose (mean ± SD = 7.88 ± 2.14 versus 9.66 ± 4.17 mmol/L; P = 0.151) and incidence of the aforementioned GI symptoms (18.2 versus 17.6%; P = 1.000) between those who harbored B. hominis and those who did not. Surgical history revealed a higher proportion of the controls with appendectomy (7.8%) compared with positive cases (0%; P = 0.010). A history of cholecystectomy was also found more frequently in controls (5.7%) than in positive cases (1%), and this difference approached statistical significance (P = 0.065).

All subjects underwent UGI endoscopic and sigmoidoscopic examination and 64 of them had gastric biopsies performed. Abnormal UGI endoscopic findings among individuals with or without B. hominis are shown in Table 2. Case-control comparisons did not reveal any relationship between any UGI endoscopic and sigmoidoscopic findings and the acquisition of B. hominis. Among those subjects who had received gastric biopsies, H. pylori was detected more frequently in the positive cases (19 of 26, 73.1%) than the controls (15 of 38, 39.5%; P = 0.017). Histologic examination of gastric biopsy specimens did not disclose B. hominis. Notably, none of the subjects discharging this organism showed appearance of colitis or colonic ulceration on sigmoidoscopic examination, or had leukocyte detected in the fecal examination. There was also no significant difference regarding the blood cell count, erythrocyte sedimentation rate (Table 3), incidence of eosinophilia (2.02% versus 4.15%; P = 0.503) and guaiac test results (14% each) between B. hominis-positive and -negative individuals. However, the results of serum chemistry examination (Table 3) showed that the concentration of urate was lower in positive cases than in controls (mean ± SD = 361.64 ± 87.44 versus 392.57 ± 93.38 μmol/L; P = 0.006), as was concentration of triglyceride (mean ± SD = 1.48 ± 0.94 versus 1.77 ± 1.33 mmol/L; P = 0.079). Of the tumor markers tested, a higher percentage of positive cases had a value of carcinoembryonic antigen greater than 6 ng/mL (the upper normal limit in this hospital) compared with controls (7.1% versus 3.6%), but this difference was not significant (P = 0.247). Interestingly, the incidence of HBsAg positivity was significantly higher in the positive cases than the controls (20.2% versus 7.3%: P = 0.002). Analysis of the 34 subjects with HBsAg positivity showed that those who harbored B. hominis tended to have alanine transaminase levels > 0.67 μat/L compared with those who did not harbor the organism (5 of 20 versus 1 of 14; P = 0.364). Among these chronic HBV-infected individuals, the incidence of GI symptoms also was not increased in those harboring B. hominis (2 of 20 versus 1 of 14). The rate of positivity of antibodies to HCV was quite low in the B. hominis positive cases, as in the controls (2.1% versus 1%; P = 0.665). Twenty-four subjects received a serologic test for HIV infection. All showed negative results. Variables that were significant (P < 0.05) or approached significance were included in the multivariate analysis. The HBsAg positivity and lower concentration of urate remained the two variables that were independently related to the acquisition of B. hominis (both P = 0.009).

DISCUSSION

Infection with B. hominis occurs worldwide, with prevalences ranging from 1.5% to 10% in developed countries and from 30% to 50% in developing countries.¹¹ The results of this study indicate that the prevalence of this protozoan in Taiwan, which has not been previously reported, is quite low (1.13%). Jelinek and others surveyed German tourists returning from different areas throughout the world and found that those returning from East Asia and North America had the lowest rate of B. hominis infection (1.1%).¹² In a recent epidemiologic survey in Japan, the incidence of this organism in apparently healthy Japanese was only 0.5%.⁷ The investigators attributed the lower prevalence of this protozoan in their country to their rigid sanitary practices. The prevalence of this parasite was significantly lower in asymptomatic than in symptomatic subjects in a previous study.¹² This finding is similar to the present study, in which individuals harboring the organism were mostly asymptomatic and apparently healthy.

Reported risk factors for acquiring B. hominis include a history of travel to the tropics, the consumption of untreated water,⁵ handling animals,¹³ abnormal GI tract function, and immunocompromised status.^11,14 Diabetes mellitus, an immunocompromised status, has been suggested to be associated with the acquisition of blastocystis in several reports.^15–17 However, these associations were based only on case studies and have not been substantiated.¹¹ In this case-control study, no association was found between diabetes mellitus and B. hominis infection. Similar to a previous report that found patients with alcoholic cirrhosis and chronic hepatitis experienced more severe symptoms when infected with B. hominis,¹⁸ we found that individuals with chronic HBV infection were at increased risk of acquiring B. hominis. Among the 34 individuals with chronic HBV infection in this study, individuals who harbored B. hominis were more likely to have an abnormal liver function test result than those who did not (5 of 20 versus 1 of 14), but this difference was not significant. Nevertheless, most of the chronic HBV-infected individuals in this study were chronic carriers as reflected by their normal liver function test. There have been reports indicating a defect of immune response in chronic HBV carriers.^19,20 Whether this immunodeficiency in HBV carriers also predisposes to the acquisition of B. hominis is unknown.

The pathogenic role of B. hominis in causing clinical symptoms remains a controversial issue.¹ In this study, most (87.5%) of the individuals harboring this organism were asymptomatic. In addition, the incidence of GI symptoms was comparable between individuals who were positive or negative for B. hominis. Thus, no association was found between B. hominis and the development of GI symptoms in this population. Even in individuals with immunocompromised status, such as those with diabetes mellitus and chronic HBV infection, the incidence of GI symptoms was not increased in those who harbored this parasite.

Whether any intestinal abnormality promotes growth of B. hominis or whether this organism can cause intestinal pathology also remains unclear. Udkow and Markell suggested that any abnormal intestinal pathology would allow for proliferation of this protozoan.² Horiki and others found that four patients who had intestinal obstruction due to cancerous growths were heavily infected with B. hominis. They suggested that the intestinal obstruction might have permitted the more abundant growth of B. hominis.²¹ In this controlled study, however, the UGI endoscopic and sigmoidoscopic examinations did not reveal any mucosal lesions that were correlated with the acquisition of B. hominis. Nevertheless, whether there is any functional GI disorder that promotes the growth of B. hominis cannot be elucidated from this study. In a recent study, Giacometti and others found that B. hominis infection was associated with irritable bowel syndrome.²² They speculated that the B. hominis was more likely to be an indicator of intestinal dysfunction or a disrupted intestinal ecosystem in these patients, but not a causative agent of this functional GI disturbance.

In a recent murine model of B. hominis infection, distention of cecum and colon were found on necropsy.²³ The histologic examination of the cecum and colon showed intense inflammatory cell infiltration, edematous lamina propria, and mucosal sloughing. However, the pathogenicity of B. hominis in human intestine has only been elucidated from case reports and a case series study.^3,4 Although there have been case reports suggesting invasiveness of this organism,²⁴ there were also reports showing that B. hominis rarely caused intestinal pathology.^5–7 In this controlled study, none of the positive cases had appearance of colitis or colonic ulceration on examination of colon up to 60 cm above the anal verge, or had leukocytes detected on stool examination. We did not examine the upper colon in this study. In a previous study, however, the colonoscopic examinations of seven subjects, some of whom passed large numbers of B. hominis, did not find any B. hominis-related intestinal lesion.⁷ This controversy in pathogenicity may result from the existence of virulent and avirulent strains of B. hominis because these protozoa were found to be antigenically and genetically heterogeneous.¹

In a small number of individuals who had received gastric biopsy, we found an association between the presence of B. hominis and the presence of H. pylori in the gastric specimen. This is the first study with a small number of cases to report a relationship between B. hominis and H. pylori. Further studies with a larger case number needed before a conclusion can be reached. Nevertheless, it is generally suggested that B. hominis is transmitted by the oral-fecal route,¹¹ as is H. pylori.²⁵ Therefore, a proportion of H. pylori may be transmitted via the same route as that of B. hominis. In some instances, certain B. hominis isolates grew faster when cultured in the presence of accompanying bacteria.¹ Furthermore, Acanthamoeba castellani, a free-living ameba, was shown to have an ability to enhance growth and survival of H. pylori when these two organisms were co-cultivated.²⁶ However, the possibility of a symbiotic relationship between H. pylori and B. hominis is quite low because H. pylori colonizes the UGI tract but B. hominis, a strictly anaerobic organism, was predominantly found in the cecum in an animal model.²⁷

In the multivariate analysis, a lower concentration of urate remained a significant factor associated with B. hominis. It had been suggested that B. hominis would become a pathogen under specific conditions, such as immunosuppression, poor nutrition, or concomitant infections.²⁸ The lower concentration of urate in our positive cases may not represent a status of poorer nutrition. Even though the B. hominis positive cases also had a lower level of serum triglyceride, the albumin level, a sensitive indicator of nutritional status,²⁹ was comparable and within normal limits in the two groups. Interestingly, hypouricemia has been found to be a poor prognostic indicator in HIV-infected patients with central nervous system infection²⁹ and in patients with intra-abdominal sepsis.³⁰ It has been suggested that decreased food intake and/or increased renal urate loss may be the cause of hypouricemia in HIV-infected patients. However, the exact relationship between hypouricemia and poorer prognosis in these patients is unclear.²⁹

In conclusion, the pathogenic role of B. hominis in causing GI symptoms and intestinal pathology could not be delineated in this study. The presence of this organism had no association with any intestinal lesions revealed by UGI endoscopy or sigmoidoscopy. However, confirmation of the findings that individuals with chronic HBV infection were at increased risk of acquiring B. hominis and the association of lower urate concentration with this organism requires further evidence.