|Values are valid only on day of printing.|
Click CC box for captions; full transcript is below.
Published: July 2010Print Record of Viewing
In Part 2, Dr. Pritt provides an overview of the molecular methods available for detection and diagnosis of tick-borne infections.
In Part 1, Dr. Matthew Binnicker discussed the conventional methods available for detection and diagnosis of tick-borne infections.
Presenters: Matt Binnicker, PhD
Presenters: Bobbi Pritt, MD
Welcome to Mayo Medical Laboratories' Hot Topics. These presentations provide short discussions of current topics and may be helpful to you in your practice.
Our speaker for this program is Dr. Bobbi Pritt, Assistant Professor of Laboratory Medicine and Pathology and Director of the Clinical Virology and Parasitology Laboratories in the Division of Clinical Microbiology at Mayo Clinic. In Part 2, Dr. Pritt provides an overview of the molecular methods available for detection and diagnosis of tick-borne infections. In Part 1, which is also posted, Dr. Matthew Binnicker discusses the conventional methods available for detection and diagnosis of tick-borne infections.
In this presentation, we’ll discuss the molecular approaches to diagnosis of tick-borne infections, with an emphasis on how PCR can be used to confirm or supplement conventional methods of diagnosis, such as serologic tests. This presentation is a follow-up to the hot topic given by Dr. Binnicker on conventional diagnostic methods of tick-borne disease.
As outlined by Dr. Binnicker in part I of this hot topic, conventional laboratory methods for the diagnosis of tick-borne infections include smear microscopy and serology.
Adjunctive or alternative methods for diagnosis include molecular diagnostics (specifically PCR), in situ hybridization (such as used on formalin-fixed, paraffin-embedded tissue), and sequencing of amplified products. The focus of my presentation today is on PCR and how it can be used in the diagnostic laboratory.
First, to review, tick-borne diseases are endemic worldwide, including both temperate and tropical climates. In the United States, major diseases include Lyme disease, anaplasmosis, and babesiosis, all which are transmitted through the bite of an Ixodes hard tick, such as shown here on the right. Lyme is actually the most common tick-borne infection in both North America and Europe, and is caused by the bacteria Borrelia burgdorferi. Ehrlichiosis, Rocky Mountain spotted fever, and tularemia are also examples of tick-borne diseases in the United States.
There are both advantages and limitations to conventional tests for these diseases.
First of all, these tests have been in use for a long time, so they are familiar, proven to provide utility in certain situations, and they are widely available in clinical laboratories. Therefore, they are used in diagnostic algorithms published by nationally and internationally recommended committees and professional societies.
However, subjective morphologic interpretation requires experience, such as interpretation of blood smears, and formation of antibody response for serologic tests typically takes 4 to 6 weeks following exposure. Finally, cross-reactivity and nonspecific serologic responses are common and may be misleading.
Therefore, molecular diagnostic methods offer potential advantages as adjunctive or alternative tests. They often have improved sensitivity and specificity compared to subjective morphologic exams, and they may be able to detect nucleic acid in acute disease, prior to the formation of a detectable serologic response. Compared to serology, they are also a better method to test for cure, since the presence of DNA or RNA will disappear prior to the antibody response, which can last for months to years after infection.
They are not without potential disadvantages however. PCR often is more expensive than conventional methods, although I must note that PCR methods can be cost effective in the long run, if labor, decreased hospital stay and less unnecessary medication are taken into consideration. PCR also requires high complexity testing facilities with skilled personnel and strict contamination controls. Because of this, PCR tests for agents causing tick borne disease may have limited availability. Finally, it is worth noting that nucleic acid from nonviable organisms may be detected after successful treatment, and it can take a week or more to clear the bloodstream. Therefore it is important to keep this in mind if PCR is being used to determine efficacy of treatment.
In order to help make some sense of how both conventional and molecular tests can be used for guiding treatment and diagnosis of tick-borne diseases, I will be referring to recommendations put forth by widely accepted organizations such as the Infectious Diseases Society of America or IDSA and theCDC. I will highlight some of the important aspects of these guidelines in the following cases.
So our first case is that of a a 42-year-old man from Rhode Island who presents with high fever and shaking chills. He is a self-professed world traveler, and he had just returned from a 2-week “safari” in Tanzania 1 month prior to his presentation. And while he was in Africa, he said that he had multiple mosquito bites, but he took his malaria prophylaxis as prescribed. He also reports tick exposure from woods behind his home.
On diagnostic workup, blood and urine cultures were obtained to exclude a systemic infection. And given his recent travel history to Africa, there was also a concern for malaria so peripheral blood smears were ordered.
Now shown here is a representative image of his Giemsa-stained peripheral blood smear, which shows multiple red blood cells with intracellular parasites mostly in the shape of rings. You’ll notice that the rings are round, although there are some oval shapes, and there is some degree of size pleomorphism present.
Based on these morphologic features, the differential diagnosis includes a potentially fatal form of malaria due to Plasmodium falciparum infection, and it also includes babesiosis. Unfortunately, these 2 organisms can be quite difficult to distinguish from one another by morphologic means.
So to demonstrate this, I’ve shown these organisms side-by-side in this image. One of these is Plasmodium falciparum and one is Babesia microti, which is the most common cause of babesiosis in the United States, and you can see that they look virtually identical. So to answer the question that I pose at the bottom of the slide, let’s review the main morphologic features of each.
First, it is important to note that neither organism enlarges the size of the red blood cell that they are infecting. And this is in contrast to infections with Plasmodium ovale and plasmodium vivax, which are other species that cause human malaria. The parasitic forms of Plasmodium falciparum and Babesia speciesare also similar to one another in that they most often form ring-shapes.
Less commonly, banana-shaped gametocytes may be seen in Plasmodium falciparum infections, which are diagnostic for this parasite. Also, the rings of the Babesia species are typically more pleomorphic than those of Plasmodium falciparum, and it’s common to see oval, spindled, or even amoeboid forms. When they align in a tetrad or what is called a maltese cross configuration, this is diagnostic for babesiosis. Finally, extracellular forms are a common feature of babesiosis but not Plasmodium falciparum.
Here is an example of the classic maltese cross. And although this is diagnostic for babesiosis, this form is rarely seen on peripheral blood smears.
This image shows a beautiful example of the extracellular forms of babesiosis. And notice how the extracellular rings appear very similar to those that are inside the red blood cells.
So now that we’ve gone over the differences between the 2 parasites, let’s go back to our side-by-side comparison and try to differentiate them.
Notice that both parasites have thin ring forms and the infected red blood cells are not enlarged compared to the neighboring uninfected red blood cells. However, note that some of the rings on the left are much more pleomorphic than those on the right, and that some of the ones on the left are distinctly oval. Therefore, a skilled microscopist can say that the smear on the left is that of Babesia species while the one on the right is Plasmodium falciparum. But, this gives you an appreciation of the subtleties required for morphologic diagnosis.
Now in this particular case, there was enough uncertainty regarding the microscopic diagnosis, especially given the patient’s recent travel history to a malaria endemic area, that PCR for babesiosis and malaria was also ordered.
The malaria PCR used in our laboratory targets a segment of 18S ribosomal DNA, and it differentiates the 4 main species by melting curve analysis. Shown here on the right are the melting peaks for Plasmodium ovale, vivax, falciparum and malariae, the 4 main species of Plasmodium that infect humans. In this case, the PCR was negative, which was a relief for the patient and the physician.
Now in contrast, the PCR for Babesia microti DNA was positive, so that confirmed the morphologic impression of babesiosis and it allowed the patient to be successfully treated.
So, I thought that this case nicely highlights some of the advantages and potential uses of PCR in this setting. First, it allowed for confirmation of the morphologic identification which was made on the conventional blood smear. And this allowed for differentiation between similar appearing organisms, given that Plasmodium falciparum was in the differential diagnosis. Another potential advantage to PCR is that it may allow for detection of low parasitemia that could be missed on a peripheral blood smear. And in general, PCR is more sensitive than peripheral blood smear for detection of intra-erythrocytic parasites.
The use of PCR also fits into the IDSA guidelines that I had mentioned previously, which state that active babesiosis should be diagnosed by the clinical presentation – in this case, viral -like symptoms AND identification of parasites in the blood by either smear or PCR. Now, because patients may have persistent antibodies to babesia from prior exposure, serology alone does not warrant treatment for babesiosis.
A potential disadvantage of Babesia PCR is that it may not be rapid enough for primary diagnosis. Malaria and babesiosis are potentially fatal diseases, so diagnosis should be performed in a STAT manner. And if PCR cannot be performed immediately, 24 hours a day, then either a peripheral blood smear or antigen detection methods must be available instead. Also, should a result be positive for Babesia or Plasmodium species, then calculation of the percent parasitemia needs to be undertaken, and this still requires a peripheral blood smear.
Another potential disadvantage of PCR in this setting is that PCR may detect residual DNA when no viable organism is present. In a study by Krause and colleagues, babesial DNA persisted for a mean of 16 days in 22 subjects following treatment, so it can take some time to clear all of the DNA from the blood stream. Also, PCR tests for babesiosis may not detect all species, and the clinician needs to be familiar with the specificity of the test used by the laboratory.
In addition to Babesia microti, there are several different species of Babesia that cause human babesiosis, including Babesia divergens, Babesia duncani (which was previously referred to as WA-1 because it was first isolated in Washington state), and the newly described MO-1 strain from Missouri State. And these organisms are genetically distinct; so they may not be detected by PCR assays that are specific for Babesia microti.
So now let’s move on to case number 2. A 4-year-old boy from Connecticut was brought to his pediatrician by his mother for a 2-day history of fever and headache. During physical exam, the pediatrician noticed a 5-cm maximum dimension, oval-shaped rash on the boy’s back that the mother hadn’t seen previously.
On questioning, she did not recall any injury or any insect bites to this area that would account for the rash. But, the physician was concerned that it may represent an atypical rash of Lyme disease, and this suspicion was supported by the fact that the patient had just returned from a camping trip in the woods 5 days ago.
Now as mentioned by Dr. Binnicker in part 1 of this hot topic, the diagnosis of Lyme disease is often clinical, and is based on symptoms and objective clinical findings such as the classic “bulls-eye rash” which is otherwise known as erythema migrans and other suspicious symptoms such as facial palsy and arthritis. Laboratory testing is generally not recommended when patients present with erythema migrans or they lack symptoms, exposure history, or they’re from a nonendemic area. But in this case, the patient does have symptoms, even though are not classic and the family does live in an endemic area.
And you can see from this CDC map showing the reported cases of Lyme disease in 2008 that Connecticut really is in the heart of the endemic Lyme area. In fact, Lyme disease was first discovered in a small town in Connecticut called Lyme, which is how the disease got its name. So, the physician, based on this history, decided to order some additional laboratory tests.
Now, as Dr. Binnicker discussed, serological assays are the test of choice for Lyme disease, and are typically performed in a tiered approach, starting with a screening enzyme immunoassay or EIA. However, as shown in this chart, the sensitivity of the EIA is dependent on the stage of disease. And notice that in stage 1 or localized disease, the sensitivity of the EIA is only 74%.
Therefore, the physician ordered both an EIA and PCR. The EIA for Borrelia burgdorferi antibodies in this case was negative, but the PCR for Lyme disease was positive, and the physician was able to diagnose and effectively treat the child. And thankfully, the patient’s symptoms resolved with doxycycline treatment.
This case nicely demonstrates some potential roles for PCR in the diagnosis of Lyme disease; specifically that it may facilitate diagnosis of early infection before a serologic response is detectable. It’s also useful for confirmation of active disease in a previously infected patient who has a positive serology, and this is actually quite important in endemic areas because it is not uncommon for individuals to have a detectable serologic response due to previous disease which may have even been asymptomatic, whereas Borrelia burgdorferi DNA would only be detected in active infection. PCR can also be useful for diagnosis of an atypical rash when tissue from a skin biopsy of the rash is tested. And it can be used for confirmation of neuroborreliosis (especially early disease), and Lyme arthritis. In fact, many different types of specimens can be tested for Lyme PCR.
The question is, which specimens should be tested and which ones provide the greatest diagnostic yield? To examine this question, we retrospectively examined the positivity rates of over 23,000 specimens tested by real-time PCR for Borrelia burgdorferi DNA. And it turned out, synovial fluid and tissue samples gave the highest yields, with positivity rates of 6.4% and 6.5% respectively, and surprisingly, CSF and blood only had positivity rates of 0.1% or less. Even more concerning is that 8 of the patients with a positive synovial fluid PCR result also had a blood specimen that was tested concurrently, and all 8 of these specimens were negative. So if only blood had been tested, these cases would have been missed.
This would support what other authors have found regarding testing of blood by PCR for Lyme disease. In general, testing of this sample type provides low clinical sensitivity, ranging from 10% to 52%. Positive results seem to correlate most with the presence of signs and symptoms of disseminated disease. And because of the low sensitivity, the CDC actually cautions against performing PCR on blood and urine for Borrellia burgdorferi DNA, and it considers them to be inappropriate specimens. Unfortunately, this isn’t widely appreciated, and many physicians continue to order Lyme PCR on blood in lieu of other more appropriate specimens or recommended serologic tests.
Now let’s now move on to a short related case. A 41-year-old male presents to the emergency room with a tick he had just removed at his home and he presents the tick to the nurse for examination who notes that it is clearly engorged with blood. According to the patient, the tick was most likely acquired on a recent hike he had taken last week, and therefore, it was probably present for about 5 days. The patient now requests that PCR be performed on the tick so that he will know whether or not he has Lyme disease. And he also asks if he can receive antibiotic prophylaxis for Lyme disease.
Here is the tick that was sent to the microbiology laboratory for identification.
Based on morphologic features such as scutal markings and festoons, the lab was able to conclude that this tick is NOT an Ixodes tick, and instead, it is most likely to be Dermacenter varabilis, the dog tick. Now, since Lyme disease is only transmitted by Ixodes ticks in the United States, the patient can be reassured that he did not acquire Lyme disease from this particular tick. However, this raises the good question of whether or not it is valid to test a tick for Borrelia burgdorferi DNA, and how tick identification may play a role in Lyme prophylaxis.
Now according to the IDSA guidelines for prophylaxis of Lyme, routine prophylaxis is not recommended after a recognized tick bite. However, a single dose of doxycycline may be offered to adult patients and children when ALL of the following circumstances exist.
First, the tick must be definitively identified as Ixodes scapularis, which is the main vector of Lyme disease in the US. The tick must have also been attached for 36 hours or more as determined by the exposure history and degree of engorgement of the tick. It must also be possible to administer prophylaxis within 72 hours of the time that the tick was removed, the local rate of infection of the ticks with Borrellia burgdorferi must be at least 20%, and doxycycline should not be contraindicated for use in the patient in question. Since this particular tick is not Ixodes scapularis, there is no indication for prophylaxis in this patient.
Regardless of whether or not treatment is administered, the IDSA also recommends that following tick removal, patients should be monitored closely for signs and symptoms of tick-borne disease for up to 30 days, including erythema migrans or a viral infection-like illness. This monitors for potential Lyme disease, as well as other diseases that are commonly transmitted by ticks such as babesiosis, ehrlichiosis, and anaplasmosis. Now of interest, these guidelines were recently reviewed, and the panel unanimously agreed that the 2006 Lyme disease guidelines were still valid in their current state.
So what about testing of the actual tick by PCR? Well, In general, testing is not recommended, since the presence of a tick is really a marker for exposure to ticks in general. The tick that is noticed could be positive for a number of diseases by PCR, but that doesn’t mean that it successfully transmitted any of them to the host. Or, the tick could be negative, but the patient could still have Lyme disease from a separate exposure – even from a tick that the patient didn’t realize was present. Therefore, it’s really best to rely on guidelines as previously mentioned for treatment decisions regarding tick-borne diseases rather than performing any test on the tick itself.
So now let’s go on to our last case, which is a 20-year-old male from Wisconsin and on the right is shown the county of his residence which you can notice is right along the Mississippi river which separates Wisconsin and Minnesota states and this is an area known to be endemic for multiple tick borne diseases. He had previously received a kidney transplant and so he was on immunosuppressive therapy. And he now presented with fever, malaise, and headache. Lab testing showed lymphopenia and mildly elevated liver function tests, and a peripheral blood smear was negative for babesia parasites. Further testing was performed, including a serologic panel for babesiosis, anaplasmosis, and Lyme disease and these were all negative. But, the physician was still concerned about an infectious etiology given that this patient was immunosuppressed, so a tick-borne disease PCR panel was also performed.
The tick borne PCR panel consisted of assays for Babesia microti which was negative, Borrelia burgdorferi which was also negative, and finally Ehrlichia and Anaplasma species and this was positive. So just as a reminder, Ehrlichia chaffeensis is the agent of human monocytic ehrlichiosis, otherwise known as HME and Anaplasma phagocytophilum causes a disease called human granulocytic anaplasmosis, otherwise known as HGA. Lastly, there is Ehrlichia ewingii which causes human ewingii ehrlichiosis and this PCR would detect all three of these organisms.
As you can see from this table, the sensitivity of various diagnostic modalities for ehrlichiosis and anaplasmosis varies with the stage of disease. Highlighted here is the most sensitive method for diagnosing human ehrlichiosis and anaplasmosis during the acute state which is PCR.
In fact, PCR is becoming the test of choice for diagnosis of acute ehrlichiosis and anaplasmosis using EDTA- or citrate-anticoagulated whole blood. In a thorough review by Dumler and colleagues, the sensitivity of PCR ranges from 60% to 90% and it is the only definitive test for Ehrlichia ewingii at this time, since serologic tests show cross-reactivity between species, and there is no commercially available test.
This is a print out of the ehrlichia and anaplasma PCR assay performed at our institution on a Roche Lightcycler instrument. And again, the use of melting curve analysis allows for differentiation of the 3 main organisms that cause human disease. So from left to right, we have Ehrlichia ewingii, Ehrlichia chaffeensis, and Anaplasma phagocytophilum. However, notice this peak here that doesn’t line up with any of the 3 organisms. This is actually the result from this patient. And, it is clearly positive, but it seems to represent a novel species of Ehrlichia.
This prompted us to perform some additional testing including sequencing of the amplified DNA from this sample and then compare it to sequences of other closely related organisms which allowed us to produce this phylogenetic tree showing that the organism is a novel ehrlichiosis agent that is most closely related to Ehrlichia muris, a mouse pathogen. We aretherefore calling this an Ehrlichia muris-like organism or EML. And this organism has never previously been identified from humans in North America. You can see that it is distinct from other causes of human ehrlichiosis and anaplasmosis, including Ehrlichia chaffeensis, Ehrlichia ewingii, and Anaplasma phagocytophilum. So this was an unexpected and very interesting discovery.
Subsequent studies of the Ehrlichia muris-like organism have now revealed infections in at least 5 patients to date, and all of them presented with symptoms similar to human ehrlichiosis due to Ehrlichia chaffeensis such as fever, myalgias, and headache. In addition, we did some experimental testing of ticks that were gathered from the area of the presumed exposures to the EML agent. And we detected EML DNA from a single group of Ixodes scapularis nymphs. At this time, it is unknown if this tick is a vector for disease, and the reservoir host is still unknown. So there is clearly much that we can learn about this new agent of ehrlichiosis in humans in North America.
As far as testing for the EML, some serologic cross-reactivity between the EML and Ehrlichia chaffeensis has been observed at least in one case but the reliability of this cross-reactivity for diagnosis of EML infection is unknown. Therefore, PCR is the test of choice for the newly recognized pathogen at this time.
It seems probable that the EML agent is transmitted by ticks, given that the EML DNA was found in Ixodes scapularis ticks, and all of the patients with EML infection had tick exposure. And in fact, here I have a photograph of one of the patient’s backyards which demonstrates that there is pretty thick foliage and undergrowth present in the backyard which would be an ideal habitat for ticks.
So, this brings me then to my last point, which is that prevention of tick-borne diseases is best accomplished by taking some simple precautions, such as wearing protective clothing, including long pants tucked into socks and boots, use of insect repellants, such as those containing DEET, or the use of permethrin-impregnated clothing, and finally, checking for ticks after potential exposures.
So in conclusion and review, conventional laboratory methods for the diagnosis of tick-borne infections include culture, direct smear examination and serology. And while these methods are widely used, they do have some limitations. Serology is generally insensitive during the acute-phase of tick-borne infections, and therefore, PCR may play an important role in this stage of infection. But of course, remember that clinical findings should always drive the laboratory tests that are ordered.
Some specific points about PCR is that blood is NOT a preferred specimen for Lyme PCR and it does seem to have a lower sensitivity than other specimens such as synovial fluid and skin tissue for detection of Borrellia burgdorferi DNA. Also, there is a new Ehrlichia species that is closely related to Ehrlichia muris and it has been described so far from humans in Minnesota and Wisconsin and right now we are referring to this as an Ehrlichia muris-like organism and physicians should be aware of the prevalence of tick-borne diseases in their region and then consider those organisms in their differential diagnosis. So with that, I’d like to conclude this presentation.