Better tests for Lyme disease

The majority of laboratory tests performed for the diagnosis of Lyme disease are based on detection of the antibody responses against Borrelia Burgdorferi in serum. As the sensitivity of such tests increase with the duration of the infection, patients early in their illness are more likely to register a false-negative result. The testing algorithm for the condition needs to be simplified - improving sensitivity in early disease, while maintaining high specificity. We speak to Dr Adriana Marques from the National Institutes of Health about how this might be accomplished.

Throughout the US, rates of Lyme disease are climbing. With over 300,000 new cases recorded every year - around triple the rate from two decades ago - the disease is now the most commonly reported vector-borne illness in the US.

As it becomes more rampant, it is also becoming more geographically widespread. While incidences are still concentrated heavily in the north-east and upper Midwest, with 95% of confirmed cases reported from just 14 states, the ticks that carry the disease are growing in number and territory. An oft-cited reason is climate change; global warming is thought to increase both the range in which the ticks can thrive, and the amount of time they have to feed.

For those at risk of infection, the need to take precautions is clear. Lyme disease, which often starts out with a 'bullseye' rash (erythema migrans) at the site of the tick bite, typically gives rise to an unpleasant array of symptoms ranging from fever and fatigue to muscle spasms, neurological problems and heart palpitations.

Following a course of antibiotics, most patients make a full recovery, especially if the infection is caught at its early stages. Unfortunately, reaching an early diagnosis is not always straightforward.

As Dr Adriana Marques of the National Institutes of Health (NIH) explains, current laboratory methods are far from infallible.
"We have made major advances since the early days when Lyme disease was first recognised, but there are many areas in which we need further improvement," she says. "As the rate of the disease increases - and the rate of differential diagnoses increases - there is a tremendous need to improve diagnostics."

Early diagnosis
Marques, who is head of the clinical studies unit at the NIH's Laboratory of Clinical Infectious Diseases, has conducted research into many aspects of the condition. A key point of focus is improving assays right from the onset of infection.

"One of the questions we're exploring is improving early diagnosis of Lyme disease and developing a point-of-care test that could be run in the emergency room," she says. "At the moment, you have to draw blood and send it out to a laboratory, but in certain situations, say when you have someone who presents with facial palsy, a point-of-care test would be very helpful."

The problem with clinical testing at the early stages is that the vast majority of lab tests are based on antibody response. Should a person become infected with Borrelia Burgdorferi, the bacterium that causes Lyme disease, it will take a few weeks before they develop enough antibodies to register on a test. This means that for newly infected patients, the test is likely to come out negative and any diagnoses will need to be based on the physician's observations.

"For patients with a possible erythema migrans rash and a compatible history of being exposed to ticks, that patient should be treated clinically; their symptoms do not require anything further to be offered therapy," Marques says. Once a few weeks have elapsed, the patient's immune response starts to fire up and the test becomes vastly more sensitive. Unfortunately, it does not necessarily become more specific. To ascertain whether a patient really is suffering from Lyme disease, or one of several differential diagnoses, the clinician will need to factor in information such as the typicality of symptoms and geographic location.

"The predictive value of any given test will depend on its sensitivity and specificity, as well as the likelihood that the patient has Lyme disease," says Marques.

"So if you have a patient who has a very low chance of having Lyme disease, and you're doing the test mostly to rule out the disease, a positive result is more likely to be a false positive than a true positive. Each subsequent positive test will increase its predictive value, and it will become much more likely that the test is a true positive."

Holy grail
Given these acknowledged limitations, the holy grail would be a direct test to check for the presence of Borrelia Burgdorferi itself, which could theoretically enable diagnosis right from the point of infection.

This, however, has yet to become a reality. Marques points out that, while two direct testing methods have been developed - culture tests and polymerase chain reaction (PCR) - both come with significant drawbacks that curtail their diagnostic applications.

"The issue with direct tests for Borrelia Burgdorferi, is there's very little of the bacteria in the body at most stages of infection," she says. "Culture is a very difficult technique, which very few laboratories can do - it requires a special medium and it takes many weeks for the bacteria to grow. So culture is a test that's important for research, but is not useful in the clinic.

"PCR is helpful for evaluating the synovial fluid in patients with Lyme arthritis, but the chances of finding Borrelia Burgdorferi, either by culture or PCR are very low, so at this point they are not really useful in clinics. Maybe with new techniques there will be something you can use to look for the bacteria itself, but right now it's very hard to find it."

For the time being, US clinics abide by guidelines set by the Centres for Disease Control and Prevention (CDC).
Since 1995, the CDC has recommended using a two-step testing process: an enzyme immunoassay to begin with, followed by the so-called 'Western blot' test in the case of a positive or indeterminate result. While there can be difficulties associated with interpretation, the process works relatively well when recommendations are closely followed.

More concerning are the speciality laboratories - often marketing themselves as providing a second opinion - that use non-validated tests. Often, these miss out the first step and skip straight to the 'Western blot'. Taken by itself, this approach can be highly unreliable. According to a recent study that examined four such labs, three of them generated false positives and one had a false positive rate of higher than 50%.

"Patients with chronic subjective symptoms without a diagnosis can be vulnerable and desperate for an answer as to the cause of their illness. Giving them a false diagnosis based on flawed testing is misleading," wrote Dr Dan Gregson, from the University of Calgary, in a commentary published in the Canadian Medical Association Journal in August.

As Marques explains, it would be a mistake to overemphasise the failings of current testing protocol as it is followed by public health providers, and therefore opt for private testing.

Sensitivity
"If you add all the numbers together you get the impression that the tests are not very sensitive," she says. "Yes, they are not very sensitive for people early in their illness who present with erythema migrans, but they work very well on patients who have had the illness for a while."

In an ideal world, Marques says, there might be different tests offered for different stages of infection, adapting the testing algorithm to provide high sensitivity and specificity no matter how recently the bacterium was contracted.

"If a patient has a rash that could be erythema migrans, you'd use one type of test, whereas a patient who presented later, the test might be different," she suggests.

Lingering symptoms
Recently, she has been funnelling her energies towards what is perhaps the trickiest diagnostic stage of all: persistent Lyme disease infection. Long after most Lyme disease sufferers are cured, a small but significant proportion continue to grapple with lingering symptoms. This condition - officially recognised as post-treatment Lyme disease syndrome (PTLDS) and more colloquially as 'chronic Lyme disease' - has been subject to much debate, with experts divided on what it is and how best to treat it.

Marques and her team have been aiming to find a biomarker for the condition, which might differentiate patients suffering with PTLDS from those who are having complaints because of something else.

"The question is, can we find the bacteria itself after antibiotic therapy, and can we correlate that with having symptoms?" he asks.
"At this point we don't have anything that could be used in clinics, but there might be something that will be able to be developed from the large screenings of biomarkers that are being done."

As Lyme disease continues to grow in prevalence, it seems that current diagnostic techniques are due for an overhaul. As Marques sees it, this might entail simplifying the algorithm into a single test or procedure, improving sensitivity for early disease detection, and creating tests that work independently of disease duration, ideally at the point of care.

When such a test will reach the clinic, only time will tell, but for researchers breaking ground in this field, the endgame has been clearly defined.