Rapid Diagnostics for Ebola in Emergency Settings
By Nahid Bhadelia
June 26, 2015
(Click here to view the original article.)
Rapid case finding and diagnosis have been cornerstones of the strategy to stem the 2014–15 Ebola epidemic. However, during most of this epidemic, timely laboratory confirmation of Ebola virus disease has been complicated by the delay between recognition of a patient with suspected Ebola virus disease in the community, transport or self-presentation of this patient to a treatment site for phlebotomy, and the time needed to transport the sample to a laboratory equipped to run molecular diagnostics safely. This delay has had devastating effects not only at the population level, by extending opportunities for transmission, but also for individual patients.
More than 50% of patients in a suspected disease ward of an Ebola treatment facility can eventually be given an alternate diagnosis. Accurate rapid diagnostic tests, particularly those that can be completed at the point of care, could move the timing and place of diagnosis closer to the community in which a patient with suspected Ebola virus disease is first seen, halting transmission chains and allowing better allocation of scarce healthcare resources. Field deployment of rapid diagnostic tests in geographically isolated areas could also identify individuals positive for Ebola in their community who choose home care with provision of the appropriate physical resources and infection control support, preventing a potentially arduous journey for treatment that might reduce chances of survival. Additionally, the phlebotomy generally required for real-time reverse transcription PCR (RT-PCR) testing adds additional staff safety risks. Point-of-care rapid diagnostic tests that can be done with a reduced volume of capillary blood can not only reduce the risk of percutaneous injury, but might also be more amenable for use in patients who are dehydrated.
In The Lancet, Jana Broadhurst and colleagues present data from a field validation study comparing the performance of the ReEBOV Antigen Rapid Test kit, which detects the VP40 matrix protein of Ebola virus, against a benchmark RT-PCR assay, done in a field reference laboratory. Blood samples were taken from 106 patients with suspected Ebola who presented to two treatment facilities in Sierra Leone between Feb 3, and Feb 20, 2015. ReEBOV was assessed against the altona test (used plasma) as a true point-of-care modality with fingerstick samples, and at the reference laboratory using whole blood (collected by venepuncture) samples. The rapid diagnostic test had impressive results, detecting all cases that were positive by altona RT-PCR in both point of-care (28 [26·7%] of 105 samples; sensitivity 100% [95% CI 87·7–100]) and reference laboratory testing (45 [16·2%] of 277 samples; sensitivity 100% [92·1–100]).
A small subset of samples that were RT-PCR-negative were deemed positive by the rapid diagnostic test (six of 77 in the point-of-care test and 18 of 232 in the reference laboratory samples), yielding specifi city of 92·2% (95% CI 83·8–97·1 and 88·0–95·3, respectively). Discordant samples were retested with a second RT-PCR assay (Trombley), showing a greater concordance between the Trombley assay and ReEBOV than between altona and ReEBOV; however, three of 18 samples concordant on rapid diagnostic test and altona were also positive by the Trombley assay, all with a cycle threshold greater than 33·5.
Herein lies the difficulty. As the authors note, how can the performance of the rapid diagnostic test be interpreted if the benchmark method itself (altona) might not be reliable, particularly at high cycle threshold values (associated with low levels of viraemia)? Additionally, the distribution of cycle threshold values in this study were only up to 26·3 (with a positive test being cycle threshold two independent operators, and all nine disagreements between the readers occurred with a positivity score of 1+, correlating with higher cycle threshold values on altona PCR. This finding suggests that practical use of the rapid diagnostic test will need careful confirmation for patients early or late in their disease course even in the cycle threshold value range reported in this cohort. Finally, the rapid diagnostic test assay needs refrigeration for storage, which the authors suggest might be available in most secondary health centres in the endemic area, but could restrict field deployment to more geographically isolated areas in which most Ebola outbreaks start and where the need is greatest.
Broadhurst and colleagues’ Article is the only reported study for the ReEBOV test, which has been approved by WHO for emergency use. The sheer logistical, staff safety, and administrative obstacles that exist for research in Ebola treatment facilities make the successful completion of this study a feat in itself. Additionally, a dearth of data is available regarding how this and other rapid Ebola diagnostics perform in the field. In view of the drop in reported cases of Ebola since January, 2015, as the outbreak has come under control, the opportunity to do further bedside testing has decreased. With such emergent issues at stake, all data need to be accounted for, leaving researchers with the dilemma of how to interpret the available information cautiously and glean what is possible to support clinical guidelines and public health policies, while further information regarding the validity of the assay in a larger cohort is gathered.
The study by Broadhurst and colleagues makes three important contributions. First, it validates the accuracy of the ReEBOV rapid diagnostic test in patients with advanced illness who are expected to have low cycle threshold values. This result suggests that the test could be used to triage this subset of patients if RT-PCR is not available, particularly in those with a high index of clinical suspicion for a diff erential diagnosis. Second, the data presented provide crucial information about the pointof-care function of this rapid diagnostic test, such as ease of use, quality of samples taken at the bedside, and the concordance with venous samples. Lastly, the results raise caution regarding the performance of the widely used altona RT-PCR assay, which the authors suggest might have underperformed because of laboratoryspecifi c technical and performance factors, such as choice of RNA extractions and amplifi cation methods. The performance of altona RT-PCR needs to be reexamined in other laboratories operating during this epidemic; the performance limitations of the altona assay were also shown in another study from Guinea. False negatives by altona RT-PCR were of patients with Ebola who were potentially falsely ruled out and returned to their community while still infectious.
Considering that more than 27, 000 cases of Ebola have been reported in the present epidemic in west Africa, existing data for new treatments and diagnostic modalities are quite sparse. Aside from capacity building for humanitarian response, the Ebola epidemic has shown the importance of support for research infrastructure in endemic countries. In the interim, the largest challenges will be in interpretation of existing data and integration of these into the continued response.