Although SFGR have been monitored in China for nearly 30 years, infection with SFGR remains widely undiagnosed, a problem that could be the cause of delays in treatment or misdiagnoses that could lead to severe disease and fatal outcomes . The WHO Collaborating Centre for Rickettsial Reference and Research (Marseille, France) regards the IFA as the “gold standard” for the serological testing of rickettsial diseases. The drawback of this method, however, is that it requires instrumentation that is not common in rural areas. In China, the availability of the non-specific Weill-Felix’s reaction is limited to high-technology settings. However, this method is no longer used as a routine screening method for rickettsial diseases in developing countries. The other drawback to serological assays is that it requires convalescent-phase serum, a requirement that effectively rules out its use for early diagnosis. Molecular diagnostic techniques based on PCR have gradually replaced tedious culture isolation methods for disease diagnosis. Modified PCR techniques, such as nested PCR and real-time PCR, do have their merits; however, these methods are complicated and require a high-precision thermal cycler that makes them impractical for the diagnosis of SFGR infections in rural areas.
In this study, we developed a sensitive LAMP assay that detects a conserved region of the ompB gene; the limit of detection of this assay was five copies per reaction or twenty-fold lower than the detection limit for the real-time PCR assay developed in our laboratory (100 copies/reaction) (unpublished data). To evaluate the LAMP assay as a tool for clinical diagnosis, we assayed 11 clinical samples by both LAMP and nested PCR. Of the 11 samples, eight tested positive by the LAMP assay with a limit of detection of five copies per reaction, while none were positive by the nested PCR although the limit of detection was 1 copy/μL. Previous studies have described that the Bst polymerase used in LAMP is less sensitive to inhibitors than Taq polymerase used in the PCR assay [16–18]. Thus, although nested PCR has been reported to be more sensitive for detecting SFGR than the LAMP assay, we show in this study that the LAMP assay is both more accurate and more sensitive than PCR for screening clinical samples. It was notable; however, that one seropositive sample was not detected by ompB LAMP. Because of these observations, we suggest that SFGR infection be confirmed by more than one method to prevent misdiagnosis. Three suspected cases of SFGR were confirmed by LAMP but failed to score as positives by IFA. We suspect that these differences may have resulted from the failure of patients to develop sufficient antibodies for individual differences to be revealed or significant antibody titer during convalescence.
When evaluating the specificity of the ompB LAMP assay, we tested 27 members of the order Rickettsiales along with several other bacterial pathogens. The results of this study showed that no false-positive amplification was observed with the heterogeneous strains using the SFGR LAMP assay and suggests that LAMP holds promise as an accurate method for differentiating between SFGR and other diseases that cause similar symptoms. The high specificity of LAMP methods has also been reported for other pathogens [16, 17, 19–21]. Compared with conventional PCR techniques that target between two and four genomic regions, LAMP methods can use between four and six primers to target six to eight regions. Consistent with the work presented here, researchers who study other pathogens have also reported the high specificity of the LAMP assay. The reproducibility of ompB LAMP is relatively high, in that both the CVi and CVo are <5%. Although confirming LAMP results by eye is important in low-technology settings and could reduce amplicon carryover leading to false-positive results, this method was less sensitive than agarose gel electrophoresis in our study (data not shown), as well as in others . This result is in contrast to the findings of Mao et al. , which demonstrated that the two methods yielded consistent results. One concern with the SYBR Green method, however, is the risk of contamination that results from opening the tubes to add reagents. We therefore consider it important to take this risk into consideration when debating how to reduce false-positive results during field testing.
In this study, we detected SFGR DNA in samples extracted from whole blood using the LAMP assay, a kit-based assay that is considered more sensitive than freeze-thaw procedures. The problem of using such a kit, however, is that it would increase the assay cost and turn-around time. In support of using freeze-heat procedures, Paris et al.  were able to detect Orientia tsutsugamushi (order Rickettsiales) in DNA samples extracted using either a kit or by freeze-thaw procedures, where 7 out of 7 and 5 out of 7 samples were confirmed by these methods, respectively. Although the kit-based method identified SFGR with a 100% success rate, we feel that the use of boiled samples in the LAMP assay should be reconsidered for greater suitability for use in rural areas.