Dientamoeba fragilis (D. fragilis) is an intestinal protozoan with unclear pathogenic potential [Reference Shasha1–Reference van Gestel, Kusters and Monkelbaan3]. D. fragilis is commonly reported in association with gastrointestinal (GI) symptoms but has also been commonly detected in asymptomatic persons [Reference Stark2,Reference Garcia4,Reference Calderaro5]. D. fragilis is frequently detected with other organisms, complicating efforts to understand its pathogenicity [Reference Calderaro5,Reference Venturini6]. The life cycle and transmission of D. fragilis are not completely understood, and multiple hypotheses exist to explain the protozoan’s presence in human GI tracts given the fragile nature of the trophozoite stage [Reference Hall7,8]. It has appropriately been called ‘a neglected protozoan’ [Reference Stark2,Reference Garcia4]. The reported prevalence of D. fragilis varies depending on geographic location, study population, and diagnostic methods [Reference Stark2–Reference Garcia4]. Additionally, the clinical presentation ranges from asymptomatic carriage to diarrhoea, abdominal pain, and peripheral eosinophilia [Reference Garcia4–Reference Venturini6]. With the increasing availability of molecular diagnostic methods, the identification of D. fragilis has been facilitated by use of both single- and multiplex polymerase chain reaction (PCR) assays, which have a significantly higher sensitivity than microscopy [Reference van Gestel, Kusters and Monkelbaan3]. The majority of recent clinical and epidemiologic studies characterizing D. fragilis have been conducted in Europe [Reference van Gestel, Kusters and Monkelbaan3,Reference Garcia4], with the most recent study in the United States (US) being a microscopy-based study published over a decade ago [Reference Chang9]. At the time of this writing, only one FDA-cleared PCR assay is available from Genetic Signatures, and this product has been used in Australia and Europe with excellent performance [Reference Gough, Ellis and Stark10]. Our primary objective was to describe the epidemiologic and clinical characteristics of PCR-diagnosed D. fragilis patients by performing a retrospective chart review at our academic medical centre located in the US.

The University of Utah has used the GI Parasite Panel by PCR developed by ARUP Laboratories since October 2014. The panel includes Cryptosporidium hominis and parvum, Cyclospora spp., Giardia, Entamoeba histolytica, and D. fragilis targets. The D. fragilis target is a conserved sequence within the 18S rRNA gene. The analytical sensitivity is approximately 16,000 copies/ml of stool (equal to approximately 200 copies per reaction). Analytical specificity was established for each of the protozoal targets against each other and 42 additional viral, bacterial, and parasitic organisms (including Entamoeba spp. and Strongyloides). In silico analysis revealed no predicted cross-reactivity with other organisms, including all formally sequenced protozoa. All specimens were frozen immediately after collection and thawed only at the time of testing. This frozen stability was shown in validation to preserve sensitivity consistent with testing fresh stool. ARUP Laboratories recommend use of the panel for individuals with chronic diarrhoea and a travel history or other relevant exposure history or those with a complicated clinic course; the decision to order the test is ultimately left to the clinician [11].

Since the GI Parasite Panel by PCR became available, 4,804 tests have been performed on patients within the University of Utah Health system. The total positivity for any target is 181 (3.8%). For our report, a case of D. fragilis was defined by a positive PCR test; a patient with multiple positive PCR results was described as one case if there was no intervening negative result. We reviewed the charts of the D. fragilis cases to abstract relevant demographic and clinical data. Study data were collected and managed using REDCap electronic data capture tools hosted at the University of Utah [Reference Harris12,Reference Harris13]. This study was deemed exempt from full review by the University of Utah IRB (IRB_00101686).

Thirty-one samples were positive for D. fragilis (0.6% positivity). Of those 31, we identified 28 unique cases of D. fragilis, detected between April 2016 and April 2024. At least one case was identified each year, except for 2021. Apart from two cases, all patients were diagnosed in the outpatient setting, with most patients evaluated and treated in primary care clinics (Table 1). Several patients were diagnosed by gastroenterology and infectious disease specialists. The two hospitalized patients had underlying conditions, and their level of acuity was likely unrelated to the D. fragilis infection. One hospitalized individual was a bone marrow transplant recipient with concern for graft-versus-host disease as a possible aetiology of their presentation and the second was a patient with septic shock in the setting of a newly diagnosed HIV infection and multiple co-infections.

Table 1. Demographic characteristics of cases

^a Destinations visited: Columbia, Japan, Madagascar, Malawi, Mexico (4), Pacific Islands, Peru (3), Philippines, Puerto Rico, Singapore, Spain, and Vietnam.

At the time of data abstraction, 25 patients had addresses in urban Utah counties and 3 were from urban counties in nearby states. The median age was 33; 17 (61%) patients were between the ages of 18 and 49 years (Table 1). Seventeen (61%) were female. Eleven (39%) individuals reported a history of recent international travel. An additional two individuals (7%) had a history of freshwater exposure in the US. Most individuals presented with persistent GI symptoms, and several with greater than 1 year of symptoms (Table 2), and most had multiple GI complaints (79%). Approximately 82% of patients reported diarrhoea. Abdominal pain (61%), nausea (46%), bloating (39%), and constipation (25%) were also common.

Table 2. Reported symptoms

^a Missing in 4 cases.

Enteric co-detections were not commonly identified. Twenty-five (89%) cases had infectious diarrhoea testing in addition to the GI Parasite Panel PCR (Table 3). One patient was also positive for astrovirus (identified by comprehensive GI pathogen PCR panel), and another individual was positive for Blastocystis (identified by stool ova and parasite testing). A third patient was newly diagnosed with HIV and was also positive for Shigella and EPEC (also identified by GI pathogen PCR panel). In the ten patients with ova and parasite (O&P) examination results, none were positive for D. fragilis. In the ten patients with CBC results, one (10%) demonstrated eosinophilia; this was the aforementioned patient with recently diagnosed HIV and Shigella and EPEC co-detections. An additional patient was evaluated due to history of persistent eosinophilia and ultimately was diagnosed with systemic mastocytosis, a likely contributor to the eosinophilia.

Table 3. Additional infectious diarrhoea testing. Additional testing was performed on 25 (89%) cases

All individuals were treated for D. fragilis. The majority were prescribed metronidazole (89%) as initial treatment. One individual was prescribed paromomycin, another individual was prescribed tinidazole due to a history of multiple rounds of metronidazole for Blastocystis treatment, and a third was treated for concomitant chlamydia infection with doxycycline. In the 25 cases with follow-up data available, symptoms improved in 13 (52%) after one round of treatment. Seven (26%) patients were retested due to persistent symptoms following treatment; only two remained positive for D. fragilis tests upon retesting (Supplemental Table 1). Four (15%) received additional rounds of treatment with either metronidazole or doxycycline; none of those who received additional rounds of treatment experienced a resolution of symptoms.

In this single-centre retrospective study of PCR-positive D. fragilis cases over a 10-year period of PCR testing availability, we found an overall test positivity rate of 0.6%. Prior prevalence estimates vary considerably based on geographic region, population studied, and diagnostic method employed [Reference Stark2–Reference Garcia4]. Our positivity rate was higher than a 2010 study of intestinal infections in the Rocky Mountain region, which found a 0.04% prevalence of D. fragilis identified using microscopy [Reference Church, Neill and Schotthoefer14] and notably lower than the reported prevalence of D. fragilis identified using PCR in symptomatic individuals in European countries and Australia [Reference Stark2,Reference Calderaro5,Reference Stark15]. Due to the limited availability of D. fragilis PCR in the US, the clinical presentation and treatment outcomes of patients with D. fragilis in the US are not well known.

Testing was requested only on symptomatic individuals; without a control group, we cannot clearly attribute D. fragilis as the cause of the symptoms. Additional viral or bacterial testing was documented on most (89%) patients. Most patients (88%) with additional testing had D. fragilis identified as a single organism. However, three had a co-detection documented and we identified alternative diagnoses through chart review in two (irritable bowel syndrome and systemic mastocytosis). The scarcity of co-detections and alternative diagnoses is a strength of our case series as these have limited the ability to understand the pathogenicity of D. fragilis [Reference Venturini6,Reference Miguel16].

The range of GI symptoms of the patients in our study was similar compared to other studies [Reference Stark2,Reference Miguel16,Reference Clemente17]. Interestingly, only 10% had eosinophilia, which differs from prior reports [Reference Stark2,Reference Calderaro5,Reference Miguel16,Reference Garg18], though only approximately one-third of patients had CBC results for evaluation. Additionally, among the one-third of cases which also had an O&P examination performed, none were positive for D. fragilis. This is not unexpected given the high sensitivity of PCR and challenging nature of direct microscopy [Reference Calderaro19].

This study may have limited generalizability due to the single centre of data collection. Additionally, all patients in our review were tested due to the presence of GI symptoms, limiting our ability to draw conclusions about the etiologic role of D. fragilis. It is possible that other underlying causes, such as IBS, may contribute to symptomatology seen in patients in whom Dientamoeba is detected. The lack of follow-up data in this retrospective study limits our assessment of treatment efficacy.

We found that among patients from the Intermountain West who were tested using a multiparasite PCR assay, the prevalence of D. fragilis was low. Case-control studies in the US could help determine the prevalence among asymptomatic persons and better describe the etiologic role of D. fragilis. The reasons for the low prevalence in this sample of US patients compared to the prevalence in Europe require further study.