Risk Factors for Intestinal Invasive Amebiasis in Japan, 2003–2009

Amebic colitis is increasing among younger men who have syphilis or HIV.

A mebiasis is caused by the protozoan Entamoeba histolytica. Each year, this disease develops worldwide in ≈40-50 million persons and causes 40,000 deaths (1,2). There are several amebic species of protozoans; E. histolytica is a pathogenic ameba that can cause invasive intestinal and extraintestinal disease (1)(2)(3). The most
Areas with high incidences of amebic infection include India, Africa, Mexico, and parts of Central and South America (1,2,9). In countries with low incidence, such as Taiwan, South Korea, and Australia, invasive amebiasis is uncommon, but reports have indicated that amebiasis is an emerging parasitic infection, particularly among men who have sex with men (MSM) (10)(11)(12)(13). Although epidemics of amebiais have not occurred in Japan, reports from 2001 indicate that invasive amebiasis is common in middle-age men, MSM and HIV-infected patients (8,14,15). In Japan, the prevalence of amebiasis has been increasing according to data from the National Epidemiologic Surveillance of Infectious Diseases (16). However, the reasons amebiasis is increasing and the actual prevalence of amebic colitis in daily clinical practice have not been fully clarifi ed. Moreover, some studies in Japan have examined risk factors, but most of these studies have reported case series or case reports without control patients (14,15,17,18).
Several studies have indicated that HIV infection is a risk for invasive amebiasis, but no consensus has been reached on this issue (10)(11)(12)19). Furthermore, some researchers have suggested that severe invasive amebiasis may develop in HIV-positive patients (20)(21)(22). Susceptibility and clinical factors differ between HIVpositive and -negative patients because of differences in immune status. However, the effect of HIV infection on these risk factors for invasive intestinal amebiasis remains unclear.
To address these issues, we clarifi ed annual changes in prevalence and risk factors for amebic colitis among persons who had undergone endoscopy. These factors were then compared between HIV-positive and HIV-negative patients.

Study Design
We retrospectively reviewed endoscopy records for 14,923 consecutive patients who underwent colonoscopy at the National Center for Global Health and Medicine (NCGM) (Tokyo, Japan) during 2003-2009. Indications for endoscopy included screening for fecal occult blood test; colorectal cancer; anemia; examinations for symptoms such as constipation, loose stool, diarrhea, hematochezia, and abdominal pain; or therapies for colorectal adenoma, early colorectal cancer, and diverticular bleeding.
We excluded patients who had not been tested for HIV infection, syphilis, or hepatitis B virus (HBV) infection.
Patients who underwent endoscopic observation only of the anorectal area and those <15 years of age were excluded. A total of 10,930 patients were selected for analysis.
NCGM has 900 beds and is the largest referral center for HIV/AIDS in Japan. Written informed consent for procedures was obtained from all patients before endoscopy and biopsy. The study protocol was approved by the ethics committee of NCGM.

Sexually Transmitted Diseases
We collected laboratory data for sexually transmitted diseases (STDs), such as HIV infection, syphilis, and HBV infection, before endoscopy. Histories of HBV infection and syphilis were defi ned as presence of antibody against as hepatitis B surface antigen and positive results in a Treponema pallidum hemagglutination test, respectively. In Japan, because only health care workers are vaccinated against hepatitis B, positive results for antibody against hepatitis B surface antigen were attributable to vaccination in few cases.
For HIV-positive patients, we determined CD4 cell counts within 1 week of endoscopy. We categorized CD4 cell counts into 4 groups: >300 cells/μL, 201-300 cells/μL, 101-200 cells/μL, and <100 cells/μL Routes of infection were determined by medical staff who questioned each patient at their fi rst visit to the hospital. Routes were classifi ed into 6 categories: homosexual, bisexual, heterosexual, drug use, untreated blood products, and unknown. We defi ned sexual preference into 2 categories: MSM and heterosexual. Patients who were not homosexual or bisexual were regarded as heterosexual.

Diagnosis of Amebic Colitis Caused by E. histolytica Infection
We performed a biopsy and aspirated intestinal fl uid from lesions endoscopically when abnormal fi ndings were seen by endoscopy. Amebic colitis was suspected on the basis of endoscopic fi ndings, such as erythema, edematous mucosa, erosions, white exudates, and ulcers ( Figure 1) (22,23). Negative results for intestinal fl uid cultures for bacterial species or acid-fast bacillus were confi rmed. Amebic colitis was defi ned as amebic trophozoites in biopsy specimens stained with both hematoxylin and eosin ( Figure 2

Routes of Amebic Infection
When amebic colitis was diagnosed, the physician asked the patient directly for information about the route of amebic infection. The physician confi rmed whether the patient had traveled in tropical areas, resided in a facility for the intellectually disabled, was a male or female commercial sex worker (CSW), or had contact with a CSW or MSM. For travel exposure, history of overseas travel in the past year was elicited. Patients to whom none of the above applied were treated as unknown.

Statistical Analysis
We assessed changes in annual prevalence by using the χ 2 test for linear trends. We summarized descriptive data for patients with and without amebic colitis. To determine risk factors for amebic colitis, we estimated the odds ratio (OR) between amebic colitis and clinical factors including age, sex, sexual preference, and history of STDs. We divided patients into 2 age groups, >50 years and <50 years. We used a multiple logistic regression model with factors that showed p<0.2 by univariate analysis. A fi nal model was then developed by backward selection of factors that showed p<0.05. The adequacy of this model was evaluated by using the Hosmer-Lemeshow goodness-of-fi t test and a receiver operating characteristic area under the curve.
We also conducted subgroup analysis concerning HIV infection. We investigated interactions between the effect of HIV infection and risk factors for amebic colitis. In HIV-positive patients, the relationship between prevalence of amebic colitis and CD4 cell counts in 4 categories was evaluated by using the χ 2 test for linear trends. All statistical  analyses were performed by using Stata version 10 software (StataCorp LP, College Station, TX, USA).

Risk Factors for Amebic Colitis
Risk factors for amebic colitis were age <50 years (OR 11.4, 95% CI 6.

Risk Factors for Amebic Colitis
Among HIV-positive patients, age <50 years, history of syphilis, and MSM status were risk factors for amebic colitis ( Table 2). Immunosuppressed status, such as CD4 cell count <100 cells/μL, was not associated with amebic colitis among HIV-positive patients ( Table 2). As CD4 cell counts decreased, the prevalence of amebic decreased (OR 0.3; p = 0.08 by trend test).
Among HIV-negative patients, age <50 years, male sex, history of HBV infection, and history of syphilis were risk factors for amebic colitis ( Table 2). No interactions were apparent between HIV infection and risk factors, such as age, sex, history of syphilis, and history of HBV infection.

Route of Amebic Infection
Among HIV-positive patients, all 31 patients with amebic infection were male (Table 3). Of these patients, 28 were MSM and 2 were male CSWs. No patients reported contact with CSWs. The route of infection was unknown for 3 patients.
Among HIV-negative patients, 2 patients were female and 21 were male. Both female patients were CSWs. Of the 21 male patients, 8 had had sexual contact with a female CSW and 7 patients were MSM (2 bisexual and 5 homosexual). The route of infection was unknown for 6 patients.

Discussion
Endoscopic examination combined with biopsy sample collection is a valuable method for confi rming suspected amebic colitis, which is often misdiagnosed as infl ammatory bowel disease or other forms of infectious colitis caused by the similarity of associated gastrointestinal symptoms (e.g., 720 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 18, No. 5, May 2012 diarrhea, hematochezia, and abdominal pain) (14,22,23). However, only a few studies have included patients who had undergone endoscopy (17,22,23). In the present study, we performed a large number of endoscopic examinations. The prevalence of patients with amebic colitis was 0.5% (54/10,930) in this 7-year study. This prevalence was far lower than results from serum prevalence studies, which have shown prevalence in children of 8.4% in Mexico (24) and 4.2% in Bangladesh (25). However, the annual prevalence of the disease showed a tendency to increase to nearly 1% in recent years, and we assume the prevalence will continue to increase in the future. In the past, amebic infection in Japan was reportedly caused by overseas travel to countries where epidemics occurred or where amebic infection was found in residents of facilities for the intellectually disabled (16,26). However, patients with these characteristics were not observed in this study. Multivariate analysis indicated that risk factors for amebic colitis in this study were male sex, age <50 years, and histories of syphilis and HIV infection.
The reason male sex was a risk factor might be related to specifi c sexual preference (8,(10)(11)(12)(13)(14)(15) because 52 male patients with amebic colitis often had contact with MSM (n = 35) or female CSWs (n = 8). In this study, MSM constituted 90% of men (OR 4.7 for patients with HIV infection), which is consistent with results of previous reports (8,(10)(11)(12)(13)(14)(15). However, HIV-negative male patients included heterosexual patients, and ≈35% of them had had contact with CSWs. We included CSWs as routes of infection for amebiasis because amebiasis among female CSWs has been reported in Japan (27). Therefore, new infection routes other than MSM, which has been considered a risk because of a diversity of sexual activities, should be considered.
Consistent with results of past reports (8,14,15), younger age was a risk factor. One possibility is that younger age represents a risk factor because younger persons are more sexually active, although this was not clarifi ed in the present study.
Histories of syphilis or HIV infection have been noted as risk factors in previous case series (7,15,28). The present study included many patients with HIV infection or history of syphilis, which supports the hypothesis that these factors increase the risk for amebic colitis.  Among STDs, HIV infection showed the highest risk ratio, a ≈16-fold increase. HIV infection has been identifi ed as a risk factor for invasive amebiasis in many studies (10)(11)(12)21), although many details of this risk remain unclear (19,29).
We presumed that compromised immune function increased the susceptibility of patients to invasive diseases. However, no relationship was seen between low CD4 cell counts and development of amebic colitis. Under existing conditions, the reason for HIV infection representing a risk factor for amebic colitis is considered the preference for oral-anal sex as a common risk factor for both infectious conditions.
We compared prevalence and risk factors between amebic colitis patients with and without HIV infection. An incidence of 0.1% (4/5,193) has been reported in studies of HIV-negative patients with positive results for occult blood in feces (17), and our results were similar. However, annual prevalence increased in 2009 (0.38%, 5/1,316) compared with 2003 (0.05%, 1/1,878), and the rate of increase was higher than that for HIV-positive patients. This result calls for careful attention in hospitals in which patients with HIV infection are not commonly encountered. In terms of risk factors, ORs for age, sex, and history of HBV infection or syphilis in our study did not vary according to HIV infection status.
Some limitations need to be considered in this study. First, Japan has not had epidemics of amebiasis, and data in this study were obtained from a metropolitan area. In addition, our hospital treats the largest number of patients with HIV infection in Japan. Second, selection bias was present because participants were patients who had undergone endoscopic examinations, which are highly likely to be performed for healthy patients. In addition, patients suspected before examination of having amebiasis might have been more likely to be actively included in the study. Third, the number of patients with amebic colitis was small; thus, the statistical power of the study might have been low. Fourth, a retrospective design was used for this investigation. With regard to HBV infection or history of syphilis, judgments had to be made for using results of serologic testing in some cases. In addition, determination of sexual preferences and overseas travel had to be based on the self-reports of patients.
In recent years, infectious diseases caused by E. histolytica and HIV have been increasing in Japan (15,16,30). HIV infection is a particularly serious problem because its incidence is consistently increasing in Japan while decreasing in western countries (30,31).
Numbers of patients with both infectious diseases studied are predicted to increase because little is known about measures to prevent infection in association with   a diversity of sexual activities. Amebic infection, in particular, is scarcely recognized as a sexually acquired infection, and improved education is needed to prevent these diseases. In Japan, measures to prevent the spread of HIV and amebic infections are urgently needed.
In conclusion, although this study was conducted at 1 center and involved retrospective analysis of a relatively small number of cases of amebic infection, the results suggest that the number of amebic colitis patients with or without HIV infection is tending to increase in Japan. Younger men with syphilis and HIV infections are at increased risk for amebic colitis. Route of infection differed slightly in that contact with CSWs was more frequent among HIV-negative patients than among HIV-positive patients. Among HIV-positive patients, homosexual intercourse, and not immunosuppressed status, seems to be a risk factor for amebic colitis.