Diagnostic methods for differentiation of Entamoeba histolytica and Entamoeba dispar in carriers: performance and clinical implications in a non-endemic setting.

Unpreserved faecal samples, suspected to contain Entamoeba histolytica/Entamoeba dispar cysts or trophozoites on the basis of microscopic examination, and serum samples from 416 patients were collected in a prospective study to determine whether stool antigen assays and detection of antibodies in serum are reliable methods to distinguish between carriers of E. histolytica and E. dispar in comparison to the reference test: real-time PCR. In 283 patients (68%) DNA of E. histolytica or E. dispar was amplified by real-time PCR: 6 patients with amoebic colitis (2%), 19 carriers of E. histolytica (6.7%), and 258 carriers of E. dispar (91.2%). In 133 patients (31%) no DNA of E. histolytica or E. dispar could be amplified in the stool samples. This patient group was used as control for the evaluation of diagnostic tests. Using real-time PCR as a reference test, the sensitivity and specificity of (1) the Entamoeba test for the diagnosis of E. histolytica/E. dispar carrier were 59% and 98%, (2) E. histolytica II for the diagnosis of E. histolytica carrier was 71% and 100%, and (3) serology for the diagnosis of E. histolytica infection was 83.3% and 95.2%, respectively. Applied to carriers that did not originate from an endemic country the sensitivity of serology for E. histolytica infection was 90% and specificity was 98.8%. In comparison to real-time PCR the performances of Entamoeba test and E. histolytica II lacked sensitivity for a reliable diagnosis of E. histolytica/E. dispar infection in a non-endemic setting. In carriers of E. histolytica/E. dispar from non-endemic countries the high specificity of serology can be used to establish the diagnosis of E. histolytica infection if antibodies are present.