July 2008
Feature Story

Karen Lusky

Acute ischemic stroke tends to trigger “what ifs” and other questions among emergency department staff when ED patients stricken with the condition are left with crippling disabilities: “What if we could have expedited the diagnosis and treatment, or what if the patient had known he or she was high risk and had come in sooner?” And “Could this have been prevented?”

All questions that emergency medicine providers may one day ask far less often as advances in stroke diagnosis, care, and prevention—which cast laboratory testing in a central role—take hold.

The most recent breakthrough involves a novel biomarker, lipoprotein-associated phospholipase A2, or Lp-PLA2, that appears to be able to flag patients with vascular inflammation and atherosclerotic plaques at risk of rupturing.

“...Lp-PLA2 determination may provide a pivotal opportunity to appropriately classify previously misclassified persons who are actually at high risk of stroke and in need of aggressive stroke intervention,” writes Philip B. Gorelick, MD, MPH, in an article published last month on Lp-PLA2 and stroke risk in an American Journal of Cardiology supplement (www.AJConline.org. vol. 101 [12A] June 16, 2008, pp. S34– S40). The Food and Drug Administration has cleared Lp-PLA2 testing as a risk marker for ischemic stroke and cardiovascular disease.

Dr. Gorelick’s article says that while epidemiologic studies have not shown low-density lipoprotein cholesterol and other lipid factors to consistently predict stroke, “Lp-PLA2 measures have shown a consistent association with stroke risk, conferring about a two-fold increase in stroke.” Dr. Gorelick is John S. Garvin professor and head of the Department of Neurology and Rehabilitation and director of the Center for Stroke Research, University of Illinois College of Medicine at Chicago.

Gary Myers, PhD, chief of the clinical chemistry branch at the Centers for Disease Control and Prevention, says the information he’s seen on Lp-PLA2 shows that while it’s a modest predictor of coronary events, “it really does stand out for predicting stroke risk.”

Lp-PLA2 is a marker of vascular inflammation produced predominantly by macrophages and is likely to be involved in initiating the early stage of the vascular inflammatory process, said Joseph McConnell, PhD, director of clinical cardiovascular laboratory medicine at Mayo Clinic, in a recent Webinar on Lp-PLA2 sponsored by the American Association for Clinical Chemistry.

Inflammation, Dr. Gorelick says, is really at the root of atherosclerosis and formation of vulnerable plaque. As for what starts the inflammation, he postulates that traditional risk factors, such as hypertension, high cholesterol, and smoking, tend to weaken the blood vessel’s endothelium or first line of defense, causing it to become dysfunctional. This sets the stage for cholesterol to be deposited and inflammation to afflict the arterial wall.

The majority of Lp-PLA2 in the bloodstream is bound to LDL cholesterol via apolipoprotein B. The latter, “in turn, transports Lp-PLA2 to the intima of the atherosclerotic lesions,” explain Salim S. Virani, MD, and Vijay Nambi, MD, in “The Role of Lipoprotein-associated Phospholipase A2 as a Marker for Atherosclerosis” (Curr Atheroscler Rep. 2007;9[2]: 97–103). Oxidation of LDL within the intima leads to the modification of phospholipids within the LDL molecule, they say.

“This oxidative modification makes the phospholipids within the LDL molecule susceptible to the action of Lp-PLA2,” the authors write. And it’s hydrolysis of these altered phospholipids by Lp-PLA2, they say, that leads to the formation of lysophosphatidylcholine and oxidized non-esterified fatty acids, two biologically active mediators of inflammation.

Dr. Gorelick says Lp-PLA2 stains very heavily in plaque that is believed to be rupture-prone and that the marker is vascular specific. “So if it’s elevated, there is a good chance that you may have a problem in the major vascular beds with inflammation and subsequent atheroma or plaque formation,” he says.

Elevated Lp-PLA2 doesn’t predict hemorrhagic stroke, but 88 percent of strokes are ischemic, so it does predict the “lion’s share,” says Richard Lanman, MD, chief medical officer for DiaDexus, San Francisco, whose PLAC test is the only FDA-approved commercial assay for Lp-PLA2 as a risk marker for stroke and cardiovascular disease.

DiaDexus began this year to market an automated turbidimetric PLAC immunoassay that runs on multiple chemistry analyzers. The company still offers an earlier version of its PLAC test, a microplated ELISA format. Medicare pays $47 for the test, according to Dr. Lanman.

Mark Alberts, MD, director of the stroke program at Northwestern Memorial Hospital, Chicago, foresees Lp-PLA2 as “eventually being part of a screening panel, just as patients are now screened for hypertension and cholesterol.”

For now, however, an expert consensus panel, composed of first author Michael H. Davidson, MD, and Drs. Alberts, Gorelick, McConnell, and others, has recommended that the biomarker be integrated into cardiovascular prevention guidelines as an “adjunct to traditional risk assessment in patients at moderate and high 10-year risk.” The panel’s recommendations are published in the June 16, 2008 American Journal of Cardiology supplement (pp. S51–S57).

DiaDexus’ Dr. Lanman says the expert consensus panel does not recommend using the Lp-PLA2 test in people with no risk factors or only one, unless the person has a serious risk factor, such as smoking or being 65 or older. But the panel does recommend its use for those at moderate risk—having two risk factors or more or in anyone with metabolic syndrome. And that’s no small number of people: The moderate-risk category encompasses 40 percent of U.S. adults, he says, while “people with metabolic syndrome constitute nearly half of the U.S. adult population.”

Combining Lp-PLA2 with C-reactive protein, or CRP, appears to be synergistic in predicting stroke. The Atherosclerosis Risk in Communities (ARIC) study, which included more than 12,000 healthy middle-aged men and women, found that persons with elevations of both CRP and Lp-PLA2 in the upper tertile had a hazard ratio of about 11 for developing stroke when compared with individuals with low levels (first tertile) of CRP and Lp-PLA2, says the Mayo Clinic’s Dr. McConnell. “This means that if you are in that group, you are 11 times more likely, on average, to have a stroke than if you are in the group with concentrations in the lowest tertile for both CRP and Lp-PLA2.”

In explaining how the markers might act in tandem to predict stroke risk, Dr. McConnell notes that CRP is an acute-phase reactant produced by the liver in response to systemic inflammation. Lp-PLA2 is not an acute-phase reactant and is believed to reflect inflammation specific to the vasculature. Thus, when Lp-PLA2 alone is elevated, Dr. McConnell hypothesizes that it could reflect a smaller degree of vascular involvement that is not sufficient to produce an inflammatory response capable of triggering CRP production. If both markers are elevated, that may indicate that a person’s vasculature is sufficiently affected to trigger a systemic inflammatory response, leading to release of CRP. Lone CRP elevations may represent a response to systemic inflammation from another source, without significant vascular inflammation.

As DiaDexus chief scientific officer Robert L. Wolfert, PhD, puts it, “Lp-PLA2 may be an early part of the inflammatory cascade related to the vasculature, and when it does its damage, CRP responds and may put the final nail in the coffin.”

The relationship among Lp-PLA2, LDL, high-density lipoprotein, and stroke risk isn’t fully understood at this point. The CDC’s Dr. Myers says about 70 percent to 80 percent of Lp-PLA2 is carried by LDL, with the remaining 20 percent of Lp-PLA2 associated with HDL.

In mice, however, says Dr. Lanman of DiaDexus, almost all Lp-PLA2 is attached to HDL, “and mice don’t get atherosclerosis.” Thus, there may be something about LDL-associated Lp-PLA2 that confers risk, or, conversely, it may be that you need high HDL-PLA2 (like mice) to be protected, he says.

“What’s curious,” says Dr. Alberts of Northwestern, “is that in some studies, the benefits of statins seem to be independent of their ability to lower cholesterol and independent of the patient’s cholesterol at baseline. So that raises the intriguing question or hypothesis that if statins aren’t working by lowering cholesterol in everyone, are they perhaps benefiting the patient by lowering Lp-PLA2?”

Statins lower Lp-PLA2, Dr. Lanman says, as do other lipid-modifying medications and exercise. Emerging evidence indicates that weight loss lowers Lp-PLA2, and there are published data showing that statins, niacin (vitamin B3), omega 3 fatty acids, fibrates, and ezetimibe all lower it, he adds. “Anything that stabilizes plaque probably stabilizes the marker.”

No published data exist on whether aspirin lowers Lp-PLA2, though aspirin therapy does lower CRP, Dr. Lanman says. “Anecdotal data suggest aspirin therapy may lower Lp-PLA2 a little but perhaps not as significantly as it does with CRP.”

While numerous observational studies show clearly that elevated Lp-PLA2 increases stroke risk, an upcoming phase three clinical trial of an Lp-PLA2 inhibitor, darapladib, may shed light on whether lowering the enzyme alone affects outcomes.

GlaxoSmithKline is poised to start discussions with regulators about the design of such a trial, says Gary Bruell, GSK’s director of product public relations for cardiovascular/metabolic research and development. In GSK’s dose-ranging darapladib trial, the drug produced “sustained inhibition of Lp-PLA2 activity in patients receiving intensive atorvastatin therapy.”

If darapladib and other direct inhibitors of Lp-PLA2 prove to help prevent ischemic stroke and heart attack, lab testing to identify the marker’s activity rather than its mass, which the current PLAC test does, may become important. DiaDexus, in fact, already makes and sells a research product being used in clinical trials to measure such activity, Dr. Wolfert says. “We are presently determining whether it makes sense to make the Lp-PLA2 activity assay available as an [in vitro diagnostic] product,” he says.

DiaDexus is also investigating the effects of measuring Lp-PLA2 on various lipid subfractions, Dr. Wolfert says. “It’s possible that the HDL-bound enzyme would be protective, analogous to HDL-cholesterol,” he says, “but that is a hypothesis we will be able to test once the measurements are in place.”

In the clinical arena, Gerald Weiss, MD, medical director of Hunter Laboratories, Campbell, Calif., predicts Lp-PLA2 may help physicians maximize their patients’ compliance with lifestyle changes and drug therapy. Hunter Laboratories brought up the automated PLAC test a couple of months ago as part of a more comprehensive cardiovascular testing program, he says, “because we recognize that the traditional lipid panels are really just the tip of the iceberg.”

In addition to getting patients to buy in to risk-reduction treatments, the marker may be used to help follow patients with known atherosclerosis to prevent stroke. Dr. Gorelick has begun to use it at the University of Illinois to monitor patients with asymptomatic carotid artery blockage to see if they are developing vascular inflammation, signaling a possible need for more aggressive treatment to stabilize their carotid artery disease.

Evidence exists that Lp-PLA2 may be an independent predictor of recurrent ischemic stroke. Dr. Lanman says there are five published prospective epidemiological studies showing that Lp-PLA2 predicts risk of ischemic stroke. The one study of risk of recurrent stroke, the Northern Manhattan Stroke Study, a multiethnic cohort, “shows that if you have had a stroke, elevated Lp-PLA2 doubles the risk of having another one. It also predicts MI and death after first stroke,” Dr. Lanman says.

Does Lp-PLA2, or a combination of the marker and CRP, have a potential role in diagnosing ischemic stroke in the emergent care setting? Dr. Lanman notes that if both Lp-PLA2 and CRP are elevated, the person is at very high risk for stroke even if he or she appears to be low risk based on traditional risk factors. “But it’s not a test that can be used to rule in and rule out stroke or MI at this time,” Dr. Lanman notes. That said, he points to a recently published study in Germany that found that in testing chest-pain patients with a multimarker panel of Lp-PLA2, NT-proBNP (N-terminal pro-brain natriuretic peptide), and whole blood choline, when all three biomarkers were low they predicted zero risk of heart attack over the next one to two months (Möckel M, et al. Clin Chim Acta. 2008; 393[2]: 103– 109). But “an ED application for Lp-PLA2 needs more study,” he says.

The hunt continues for acute stroke biomarkers to help identify patients who may benefit from intravenous tissue plasminogen activator, or tPA, in an emergent care setting. The clot-busting therapy is usually given within three hours after stroke symptoms appear, a window of time during which the treatment’s benefits are believed to generally outweigh the risk of hemorrhage. “Clot retrieval devices are sometimes used after that time-frame,” says Daniel Laskowitz, MD, director of Neurovascular Laboratories at Duke University Medical Center. And “some hospitals have protocols where they give intra-arterial thrombolytics beyond the three-hour window in selected patients,” he says.

The problem in expediting stroke diagnosis is that, unlike myocardial infarction, which tends to present in a relatively stereotypical way, stroke may have multiple different presentations, says Dr. Laskowitz. And a number of “stroke mimics” compound the confusion. They include hypogly­ce­mia, migraine, seizure, and infectious and metabolic problems, all of which can be difficult to differentiate initially from stroke.

Hemorrhagic stroke, a major contraindication for clot-busting therapy, shows up immediately on the CT brain scan, Dr. Laskowitz says. But minimal changes can be seen on a CT scan within the first few hours after ischemic stroke. “Thus, CT is used more to rule out other sources of acute neurological deficit, like bleeding, or mass lesion,” he says.

Some research shows that multi­modal MRIs can detect ischemic stroke early, he says. “But that technology isn’t going to be available 24/7 in most settings, and you aren’t going to have someone to interpret the MRI within minutes.”

Thus, Dr. Laskowitz says, the idea is to have a point-of-care test that provides adjunctive information to identify high-risk patients who might be routed by paramedics to a stroke center or be prioritized for immediate CT scanning. “Additional testing would identify patients likely to deteriorate and might guide early in-hospital management decisions.” Having such testing available, he says, “would just be huge.”

The largest national research effort to date to develop a stroke biomarker panel that could be used at the point of care was sponsored by Biosite (now owned by Inverness Medical Innovations). The panel included brain natriuretic peptide (BNP), D-dimer, S100B, and matrix metalloproteinase-9 (MMP-9). Duke University tested the biomarker panel in a multicentered clinical trial conducted in concert with Biosite. Findings from the trial will be published soon in the American Heart Association’s Stroke, Dr. Laskowitz says.

In explaining the selection of tests for the biomarker panel, Dr. Laskowitz says BNP was first defined as coming from the brain—“it was found in pig brain.” But cardiologists “co-opted” the marker because it’s released in large quantities in congestive heart failure, he says. Yet BNP is also released in different forms of acute brain hemorrhage, including subarachnoid hemorrhage. And “BNP, it turns out, is upregulated in a robust way immediately after both ischemic and hemorrhagic stroke.”

D-dimer is an acute-phase reactant used in evaluating patients with shortness of breath who may have pulmonary embolism. When elevated, D-dimer may also indicate that someone has an intravascular clot. “D-dimer may be a marker for inflammation as well,” Dr. Laskowitz says.

In stroke, S100B is released by activated astrocytes, which are one type of glial cells, he says. MMP-9 is upregulated in inflammation and seems to be very elevated with brain hemorrhage. “Preliminary data suggest it might predict brain hemorrhage,” Dr. Lasko­witz says.

The negative predictive value for the acute stroke biomarker panel they tested was 90 percent. “That is, if a patient had a negative test, 90 percent of the time he did not have ischemic or hemorrhagic stroke,” Dr. Laskowitz says. But the test was made sensitive at the expense of its specificity, he adds, noting that making a test that’s both sensitive and specific enough for stroke is a “hard bar” to achieve in diagnosing stroke. The neurons of the complex brain compose a minority of its cells. “And a lot of different cells are involved in a very complex ischemia cascade. And whatever proteins are released [in that cascade] have to cross the blood brain barrier to get into the bloodstream” to be measured.

There was no gold standard against which to compare how the panel performed, he says. “Studies for heart injury utilize EKG and troponin [but] in the acute diagnosis of stroke, we are left primarily with clinical impression.”

Dr. Laskowitz says the “verdict is still out” on whether the stroke panel is good enough for clinical use. In his view, “Biosite was on the right track—it was a big idea and the right idea.”

Alan Wu, PhD, professor of laboratory medicine at the University of California at San Francisco, who was also involved in testing the Biosite stroke panel, views the panel as “a reasonable starting point for entering the field.” When the panel result was very high or very low, it predicted stroke or a rule out. Dr. Wu says, “The problem has been the gray zone where intermediate values were not definitive, but in my mind, other lab tests also have a gray zone, and like any lab test, [the panel] requires correlation to a patient’s clinical presentation and symptoms.”

Despite the obstacles, work continues to identify acute stroke biomarkers. Dr. Laskowitz and his team at Duke, for example, continue to collect blood samples from patients until they find a sponsor with which to work. Ultimately, they would like to evaluate a number of markers of inflammation, astrocyte activation, and neuronal injury, he says.

Now under development by CIS Biotech, Atlanta, the biomarker N-methyl-D-aspartate, or NMDA, shows promise of providing a way to help identify and potentially distinguish acute ischemic stroke from transient ischemic attacks.

As Robert Christenson, PhD, professor of pathology at the University of Maryland School of Medicine, Baltimore, explains, NMDA is “a brain receptor that’s released into the blood when endothelial cells are damaged or die resulting in increased permeability.” Of course, that could happen with brain injuries other than stroke, says Dr. Christenson, who presented on emerging stroke biomarkers during another recent AACC-sponsored Webinar. But it’s possible that even minor acute cerebral injury could cause NMDA receptor fragments to escape into the blood, triggering antibody formation against NMDA in people who had clinically silent TIAs or ones that weren’t serious enough to prompt them to go to the emergency department, he says.

Thus, CIS Biotech is developing a test for the protein NMDA itself, which would be expected to rise sharply in stroke. And it’s developing a test for autoantibodies against NMDA, which take several days at least to develop, Dr. Christenson says.

Dr. Christenson predicts it will be at least two to three years before a diagnostic biomarker is available for acute stroke. Dr. Wu foresees a panel of biomarkers for acute stroke but doesn’t think anyone will find a “magic bullet,” or as he puts it, “a troponin for stroke.”

Regardless of whether scientists produce an acute stroke biomarker or panel, the lab plays a pivotal role right now in expediting laboratory tests needed in suspected stroke patients who qualify for clot-busting and reperfusion therapy to ensure they get it in time, a point Dr. Christenson emphasized in his AACC Webinar. Only a small percent of individuals with ischemic stroke who would benefit from tPA receive the therapy known to improve outcomes. So a laboratory test or strategy that increases the number of patients could have a direct and substantial impact on patient outcomes. And numerous primary stroke centers have made great progress in reducing lab turnaround times for such testing.

In fact, Deaconess Hospital in Evansville, Ind., a Joint-Commission-certified primary stroke center, operates by the philosophy that every minute it can whittle off turnaround times for its stroke panel matters. The panel, which the triage nurse places as a standing order for patients with stroke symptoms, includes a complete blood count, a comprehensive metabolic profile, and prothrombin time and partial thromboplastin time for coagulation, says Phillip Gamble, lab manager at Deaconess. Physicians also have the option of ordering additional tests—for cardiac ischemia, for example.

Initially, the lab aimed for a 45-minute TAT from the time of order to delivery of results. And it found it could hit that mark only if everything went off without a hitch. So it took a look at how it could give itself more time.

One thing slowing down the TAT was a single centrifuge in the coagulation lab, which does the PT/PTT testing for the stroke panel. So the lab added a stat centrifuge, reducing the TAT by up to five minutes.

The lab also implemented chemistry testing using a handheld device for BUN, glucose, sodium, potassium, chloride, CO2, calcium, and creatinine to help physicians rule out stroke mimics as quickly as possible. “The lab is contiguous to the ED, so the transport time [for specimens] is minimal,” Gamble says. “Our philosophy is that the lab should do what it does best, which is to perform the test, and the ED nurses should do what they do best, which is to care for the patient.”

The POC device provides results in about a third of the time as a general chemistry analyzer, Gamble says. The results for these tests are reported in the lab information system as soon as they are ready.

“When we instituted the new POCT protocol, which has been in place for one year, we saw a difference of about 15 minutes in the TAT for the chemistry tests. We went from a median TAT of about 40 minutes from order to results to 25 minutes.”

While the faster turnaround helps physicians rule out major stroke mimics more quickly, the cost to test on the handheld devices is about eight times higher than the cost to run the tests on chemistry analyzers. And the lab initially worried that the ED physicians would demand the lab do all testing on the handheld devices. “They haven’t,” says Gamble.

On the preanalytical end, the lab also sliced off a few minutes from the centrifugation times for the stroke panel tests. And the lab is providing additional training for patient-care techs in the ED to ensure they can collect the specimen on the first attempt.

Other strategies used by Deaconess and other primary stroke centers to expedite lab test turnaround for stroke patients include:

Patient-care techs do the phlebotomy in the ED, but the lab monitors the quality of the specimens received and has cut the rate of hemolyzed specimens from 10 percent to two percent through training, Heidt says.

Patients with stroke symptoms at Alexian Brothers Medical Center, Elk Grove Village, Ill., receive a noncontrast CT before having their blood drawn for lab work, but the phlebotomist meets the patient in the imaging suite to do that immediately after the CT, says Wende Fedder, RN, MBA, director of the stroke center at Alexian Brothers Hospital Network. (Dr. Gorelick is the center’s medical director.)

At Deaconess Hospital, the lab monitors TATs for the stroke panel and a quality assurance nurse in the ED reviews stroke patients’ charts. And if the nurse finds lab results weren’t back in 45 minutes, she discusses it with the department in which the process broke down—and, in some cases, with the individual, says Lindsey Morris, RN, CCRN, CNRN, Deaconess Hospital stroke program coordinator.

Of course, not all strokes can be prevented or treated in time to stave off brain damage. And the good news is that research to find ways to help the brain recover from stroke is no longer a “sleepy area,” says University of Illinois’ Dr. Gorelick. Studies are underway using animal stroke models, he says, where libraries of drugs are being tested to see what may work, some of which may be genetically linked. Neurotrophic growth factors are also being studied.

“We need not only prevent stroke, which is exceedingly important, and treat it if it occurs, but also heighten rehabilitation methods to improve outcomes,” he says.