Modified two-tiered testing for Lyme disease
Expires: October 6, 2024
Elitza Theel, Ph.D.
Professor of Laboratory Medicine and Pathology
Director, Infectious Disease Serology Laboratory
Mayo Clinic, Rochester, Minnesota
Good afternoon everyone. My name is Elli Theel and I direct the Infectious Diseases Serology Laboratory and co-direct the Vector-Borne Diseases Service Line at Mayo Clinic and Mayo Clinic Laboratories. Today I’m going to talk about the modified two-tiered testing algorithm for Lyme disease which received CDC backing a year or so ago now, and there are also two assays that have received FDA clearance or approval for use as part of a two-tiered testing algorithm, and we’ll talk about those today.
First, my relevant disclosures are listed here.
Alright, so taking a step back, since 1994, Lyme diagnostics using serologic assays have been based on the standard two-tiered testing algorithm, which typically starts with a sensitive EIA, which, if reactive, reflexes to immunoblot testing for detection of both IgM and IgG-class antibodies to various different and well-characterized Borrelia antigens.
This algorithm, though, suffers from a number of limitations, including the issue of low sensitivity, particularly within the first 1 to 2 weeks after infection, and low sensitivity meaning it may detect fewer true positives. And that sensitivity again in that early stage ranges from about 35% to 60% depending on the study and is primarily driven by the lower sensitivity of the immunoblot assays in general. Also challenging is how immunoblots are interpreted, whether bands are visualized as present or absent by a technologist or an instrument, and there have also been concerns over the specificity of some of the included Borrelia antigens on those blots.
Despite these concerns, the standard algorithm was the best option that we had at the time and for the last 25 years, until July 2019, when the FDA cleared four enzyme immunosorbent assays (EIA) from Zeus Scientific for use as part of a modified two-tiered testing algorithm, and this has really been a historic event in the Lyme diagnostics field.
So, when it comes to the modified algorithm, the most important takeaway is that immunoblots are not used at all. Instead, the algorithm is entirely based on enzyme immunoassays or EIAs. The modified algorithm still has two tiers, but they both involve EIAs, which importantly differ in the Borrelia antigens that are used to make them.
So, currently there are actually two different versions of this modified algorithm that are FDA cleared, but regardless of the version that’s used, they both start with a first tier EIA that is based on the VlsE/pepC10 recombinant Borrelia antigens and detects total IgM and IgG-class antibodies, without differentiating between them.
If the first tier EIA result is negative, no further testing is needed. If the result is positive or equivocal, however, the second tier EIA is performed, which for both versions of the algorithm is based on whole cell sonicate material from cultured Borrelia spirochetes.
The only difference between the two versions is that in the first version, two whole cell sonicate (WCS) EIAs are done to distinguish between IgM and IgG class antibodies, versus in version two where the second tier EIA detects total antibodies without differentiating between the antibody classes. Ultimately, though, if either of the second tier EIAs are positive, the final patient result is considered positive.
With respect to performance characteristics compared to the standard algorithm, modified algorithms provide higher sensitivity for detection of anti-Borrelia antibodies, particularly during early infection.
As you can see here, in patients with erythema migrans, multiple studies have shown that across different modified algorithms, they are significantly more sensitive, by about 10 to 20%, compared to the standard algorithm in acutely infected patients. And what this means is that we can detect positive results sooner using the modified algorithm as compared to the standard testing algorithm. Sensitivity at later stages in the disease, on the other hand, is about equivalent between both algorithms. And because of that increased sensitivity during early infection, though, the overall sensitivity is ultimately higher for the modified algorithm.
With respect to specificity, the-false positive rate for the modified algorithm is roughly equivalent to the standard algorithm, and importantly, the modified algorithm remains susceptible to the same causes of cross-reactivity as is the standard algorithm, including in patients with EBV mononucleosis, syphilis, and multiple sclerosis amongst others.
Although published studies documenting performance of the Zeus modified algorithms are still pending, this is a summary of the available clinical trials data, and as you can see, they found a similar trend of higher sensitivity during acute and early Lyme disease for both versions of the modified algorithm, with identical sensitivity in later stages of infection, and importantly, and found similar specificity as compared to the standard algorithm.
Ultimately, based on these and prior studies, I would say that it is probably a guarantee that we’ll be seeing more and more EIAs and other methods that receive FDA clearance for use as part of a modified algorithm in the future.
In addition to increased sensitivity, the modified algorithm offers a number of other advantages, including allowing for detection of antibodies to a wider range of Lyme-causing Borrelia species, because the immunoblots that we were using as part of the standard algorithm were pretty specific for detection of antibodies to Borellia burgdorferi alone.
The modified algorithm may also be somewhat less expensive than the standard algorithm, and it allows for smaller labs to be able to perform both tests on-site instead of sending the testing to reference labs, resulting in a faster turnaround time for their results.
And finally, I anticipate that in general there will be less confusion around result interpretation as there will no longer be blot banding patterns to report, which caused both clinician and patient anxiety in the past.
With respect to disadvantages, remember that sensitivity of the modified algorithm is still imperfect during early disease, and a negative result still can’t be used to rule out infection. Also, given that we now only have qualitative results, we can’t monitor IgG antibody expansion as we were able to by looking at the antibody banding patterns on the immunoblot. Given that there are multiple different modified algorithm versions, there may be some confusion around how to use or how to interpret the different options. And finally, similar to the standard algorithm, results from the modified versions really can’t be used to monitor response to therapy, and it will still remain challenging to diagnosis re-infection, given these assays will remain positive for months to years following infection.
And with that, I’ll stop. Thank you for your attention.
Contact us: firstname.lastname@example.org
Photo credit: iStock.com/abadonian