Amy Carpenter Aquino
June 2021—In an AACC virtual session last December, Catherine P. M. Hayward, MD, PhD, of McMaster University, set out the stepwise approach to testing for platelet function disorders, explained the methods used to assess platelet aggregation response, and reported what most clinical labs do.
“Unfortunately, there’s no good, simple, cheap screening test” for these disorders, she said. “And to date there are some disorders we don’t really understand.”
Dr. Hayward, who is a hematologist and head of coagulation, Hamilton Regional Laboratory Medicine Program, Ontario, and professor of pathology and molecular medicine/medicine, shared the case of a 24-year-old female with bruises (many without recollection of trauma, some very large). The patient had prolonged bleeding from cuts lasting hours, delayed wound healing, prolonged bleeding after loss of her primary teeth, and heavy menstrual periods since menarche requiring medical therapy and treatment for iron deficiency anemia. No one else in the woman’s family had a bleeding disorder, and she had normal von Willebrand factor levels, Dr. Hayward said.
Her platelet count was normal, and aggregation tests showed normal response to arachidonic acid, the higher concentration of collagen, and ristocetin. “But she had quite identifiable abnormalities with other agonists, including ADP [adenosine diphosphate], the lower concentration of collagen, and thromboxane analogue. She also didn’t have secondary aggregation with epinephrine,” Dr. Hayward said.
Separate testing for platelet dense granule deficiency was negative, and the patient’s aggregation abnormalities were confirmed on another sample.
After the patient’s platelet function disorder was diagnosed based on aggregation testing, she was treated with birth control pills and, later, tranexamic acid to manage heavy menstrual bleeding while attempting pregnancy. Her iron deficiency was corrected.
The patient’s subsequent hemostatic challenges (surgery, dental extractions, and childbirth) were managed with desmopressin prophylaxis. “She achieved excellent control of hemostasis” and participated in a research study of platelet function disorders that included exome sequencing to identify causes, Dr. Hayward said.
“To date, the molecular cause of her particular platelet function disorder remains unknown.”
Some platelet disorders are autosomal dominant, and many gene mutations are known to cause platelet function problems, inherited thrombocytopenias, or in some cases both.
“Our best estimates of the prevalence of these disorders are about one to six per thousand individuals, though a population-based study has never been undertaken,” Dr. Hayward said. “Some forms of platelet function disorders, such as dense granule deficiency, are almost as common as von Willebrand disease.”
Clinical features of platelet function disorders vary. In a person with mild thrombocytopenia, there could be none. “But more typically, platelet function disorders present with mucocutaneous bleeding problems,” and gastrointestinal tract bleeding can be seen as well. “And in most platelet disorders, the bleeding typically begins soon after the challenge,” with the exception of Quebec platelet disorder, Dr. Hayward said.
A study of 1,098 adult and pediatric subjects, 482 of whom had diagnosed inherited platelet function disorders, found higher bleeding scores in the subjects with inherited platelet function disorder (median 9) than in patients with type 1 von Willebrand disease (median 5). It also found that “the bleeding scores in people with thrombocytopenic platelet disorders or platelet function disorders are higher in women,” Dr. Hayward said (Gresele P, et al. J Thromb Haemost. 2020;18[3]:732–739).
A study of the disease-causing genes responsible for bleeding, thrombotic, and platelet disorders “indicated that of the 91 genes implicated in such conditions, 61 genes are involved in disorders of platelet function and/or formation,” she said. “And among these 61 genes, almost half of them are conditions that show a dominant inheritance” (Megy K, et al. J Thromb Haemost. 2019;17[8]:1253–1260).
“So platelet disorders are important causes of bleeding,” Dr. Hayward said, noting there are many known causes of such disorders and how they affect platelet formation or function varies depending on the type of disorder.
[dropcap]T[/dropcap]he 2015 ISTH guidance recommends stepwise testing for inherited platelet function disorders, typically done after exclusion of von Willebrand disease (Gresele P, et al. J Thromb Haemost. 2015;13[2]:314–322).
The first step is to assess the platelet count and size and the morphology by light microscopy, and to assess platelet function by light transmission aggregometry (LTA) with platelet-rich plasma—using a limited but important set of agonists and extending the agonist testing if abnormalities are found. Testing for defects in the release of alpha and dense granule contents should be done at this step or the next, Dr. Hayward said.
The second step calls for LTA with more agonists and an evaluation of secretion if not done at the first step, and additional investigations, such as flow cytometry, electron microscopy, and biochemical assays for granule contents.
“As a third step, genetic tests can be helpful for some conditions,” she said.
Aggregometry is the key test to assess platelet aggregation responses, and it can be done with platelet-rich plasma, typically by optical or light transmission. “But you can also test platelet-rich plasma by electrical impedance methods,” Dr. Hayward said. The LTA method relies on the fact that platelet-rich plasma is cloudy, and as platelets aggregate, the light transmission through a sample increases.
Whole blood aggregometry follows the electrical impedance changes in a sample, she said. When platelets stick to the electrode, the impedance changes. “It looks similar in tracings to LTA but it’s not identical,” she said. “And with collagen, both platelets and white blood cells stick to the electrode.” There are fewer publications on whole blood aggregometry usefulness compared with LTA, and one type of whole blood aggregometry using the multiplate device has been shown to be inferior to LTA for diagnosing PFDs, Dr. Hayward said.
The North American Specialized Coagulation Laboratory Association guideline published in 2010 is the only guideline that addresses how to interpret the results of aggregation tests (Hayward CPM, et al. Am J Clin Pathol. 2010;134[6]:955–963). And the ISTH guidance document on the platelet function analyzer closure times (PFA-100 is considered optional because it is not sufficiently sensitive to screen for the more common types of PFDs) and the evaluation of platelet disorders “still stands,” she said (Hayward CPM, et al. J Thromb Haemost. 2006;4[2]:312–319). (A listing of all relevant guidelines can be found at https://www.islh.org/web/index.php.)
[dropcap]”G[/dropcap]iven the recommendations that have been made on how to test for platelet function disorders, what do most clinical labs do?” Dr. Hayward asked. She reported results from an international survey of laboratories that participated in the quality assurance challenges of NASCOLA and the ECAT Foundation (Hayward CPM, et al. Int J Lab Hematol. 2019;41[suppl 1]:26–32).
“Most labs that do testing for platelet function disorders—if they offer any test—it is commonly light transmission platelet aggregometry,” Dr. Hayward said—94 percent of the 47 NASCOLA responses and 80 percent of the 61 ECAT responses. Fewer labs do testing for defects of dense granule release (32 percent of NASCOLA responses, 20 percent of ECAT responses) or alpha granule release (six percent of NASCOLA responses, eight percent of ECAT responses). Only a few labs reported doing assessments by flow cytometry or other investigations for platelet disorders.
“If we look at our own experiences, including aggregometry as part of initial bleeding disorder assessments for patients seen at our center, there are some important observations,” Dr. Hayward said. “First, like other tests that we use in coagulation labs, LTA has very good specificity”—98 percent with two or more abnormalities (Hayward CPM, et al. Semin Thromb Hemost. 2012;38[7]:742–752). LTA was performed and interpreted in accordance with the 2010 North American consensus guidelines.
“And, interestingly, because platelet disorders are more common than some other types of bleeding problems, LTA yields a fairly good sensitivity for detecting bleeding problems”—25.7 percent with two or more abnormalities.
“If we do light transmission platelet aggregometry, lumiaggregometry tests of platelet-dense granule release, or platelet electron microscopy by whole mount methods to assess for dense granule deficiency, we generate data that can be predictive of a bleeding disorder,” Dr. Hayward said. In the case of LTA, “there’s quite a high likelihood, if the person has multiple aggregation abnormalities, that they do have a bleeding disorder” (Castilloux JF, et al. Thromb Haemost. 2011;106[4]:675–682).
With lumiaggregometry, “the odds ratio of a bleeding disorder crosses one, so this test in our hands is useful for phenotyping but not for diagnostic purposes,” she said (Badin MS, et al. Int J Lab Hematol. 2016;38[6]:648–657).
Platelet whole mount electron microscopy, which assesses for dense granule deficiency, is predictive of a bleeding problem but detects only one type of platelet disorder, she said (Brunet JG, et al. Int J Lab Hematol. 2018;40[4]:400–407).
“When we look at the variation and end points for agonist responses assessed by LTA compared to dense granule ATP release, assessed by lumiaggregometry, the CVs are much tighter for LTA end points,” she said (Hayward CPM, et al. Int J Lab Hematol. 2019;41[suppl 1]:26–32).
“Of course, with a lower concentration of ristocetin, very little aggregation is normal, so the CV for LTA with that agonist is quite high.”
With ATP release, “the CVs are above 20 with the agonists that we test,” making it more difficult to use the test for diagnostic purposes, she said.
Dr. Hayward’s group in 2018 reported its experience with electron microscopy testing for platelet function disorders due to dense granule deficiency. They count dense granules in 30 to 50 platelets to get an average count per platelet (normal range: 4.9 to 10 dense granules per platelet) (Brunet JG, et al. Int J Lab Hematol. 2018;40[4]:400–407). The study found this test to have an acceptable CV, she said. “If you confirm dense granule deficiency on another sample, then the likelihood of a bleeding disorder is quite high.” Their estimate of the odds ratio was 97 (95 percent CI, 5.4 to 1,740, P<0.01).
The respective sensitivity, though, of LTA and ATP release for detecting dense granule deficiency indicates “we can’t use those other tests to screen for dense granule deficiency,” she said, “as we only picked up multiple agonist abnormalities by LTA in 52 percent of our confirmed dense granule deficient cases, and ATP release tests only picked up 70 percent of the cases.”
If platelets are tested with a strong agonist, aggregation, thromboxane generation, and secretion will be induced at the same time, Dr. Hayward said. “And that affects what you see if you’re monitoring a tracing on lumiaggregometry, for example. If you test weak agonist, feedback is needed. So thromboxane synthesis needs to happen, secretion needs to happen, in order to get the maximal aggregation with a weak agonist” (Cattaneo M. Semin Thromb Hemost. 2009;35[2]:158–167).
Secretion will not be seen at the same time as aggregation with a weak agonist, she said, because the positive feedback loops need to affect platelet function before there is activation to the point of granule secretion.
[dropcap]L[/dropcap]ooking at LTA findings for native (undiluted) versus platelet count adjusted platelet-rich plasma samples for healthy controls, “there are very similar findings with most agonists, but if you look at weak agonists, such as lower concentration of ADP, and epinephrine, the findings differ,” Dr. Hayward said.
“So it’s important that when you’re doing this test for clinical purpose, you validate what the reference intervals are for the procedure as you do it,” she added. “With ristocetin, the lower concentration is more variable with native samples.”
When her group analyzed the superiority or inferiority of the two methods, they identified that using platelet count adjusted platelet-rich plasma for the purpose of detecting a bleeding disorder by LTA was superior to using native platelet-rich plasma (Castilloux JF, et al. Thromb Haemost. 2011;106[4]:675–682).
They analyzed which of the aggregation agonists they were using in their panel were the most helpful for diagnosing platelet disorders, and they found that a few agonists detect most platelet function abnormalities (epinephrine, arachidonic acid, thromboxane analogue, and 1.25 μg/mL of Horm collagen) (Hayward CPM, et al. J Thromb Haemost. 2009;7[4]:676–684). “So I think it’s important to include these agonists in your testing,” she said, but “we also need to choose agonists that allow us to detect and distinguish rare disorders and disorders that have distinct causes.”
To detect P2Y12 defects—a problem with the ADP receptor—“we need to include ADP in our testing,” she said, noting that P2Y12 deficiency affects not only the platelet responses to ADP but also agonist responses dependent on ADP feedback.
To detect aggregation abnormalities due to Bernard Soulier syndrome or von Willebrand disease, “we need to test with ristocetin.” Testing with collagen is necessary to detect collagen receptor defects (e.g. those that affect glycoprotein VI).
And to detect thromboxane generation and response defects, and distinguish those conditions from other types of platelet function disorders, “we need to include both arachidonic acid and thromboxane analogue in our test panels,” she said.
Dr. Hayward’s recommended aggregation agonist panel would include ADP, collagen at several concentrations, arachidonic acid, thromboxane analogue U46619, ristocetin at high and low concentrations for detecting loss- and gain-of-function problems, and epinephrine “because this agonist turns out to be useful for detecting common platelet function disorders.”
When evaluating aggregation results, she advises first questioning whether the LTA testing was done in accordance with guidelines, and whether the sample that was evaluated had a usual platelet count for the procedure.
“It’s next important to ask whether any maximal aggregation findings fell outside the reference interval,” she said, “and for the most part we’re looking for reduced maximal aggregation, with the exception being the lower concentration of ristocetin.”
If the answer is yes, but with only one agonist, “it’s possible that could be a true positive, if the patient, say, had von Willebrand disease or Bernard Soulier syndrome,” Dr. Hayward said. “If we’re thinking of the possibility of Bernard Soulier syndrome, we’d also want to look at whether the patient is thrombocytopenic and has very large platelets.”
If an abnormality is seen only with collagen, it’s possible there could be a collagen receptor defect, though a false-positive is far more likely, she said. Other single agonist abnormalities are typically a false-positive.
“If we see abnormalities with two or more agonists, then we need to go through a checklist to try to identify what the pattern suggests,” she said.
The hallmark features of aspirin defects, for example, are impaired aggregation with arachidonic acid yet normal responses to thromboxane analogue. For Glanzmann thrombasthenia, the hallmark features are impaired or absent aggregation with all agonists and evidence of an agglutination but no aggregation response to ristocetin. With P2Y12 receptor defects, “we’re looking for a very striking defect in the ADP response that can be accompanied by other agonist abnormalities.”
“The most common thing to see, though, is another type of pattern that is a typical feature of the more common platelet disorders that have been more broadly classified as secretion defects,” she said.
If all of the maximal aggregation findings are within the reference interval, then the findings are nondiagnostic. “Other accompanying abnormalities, such as deaggregation, usually accompany reduced maximal aggregation,” she said.
Also important is to consider the evidence if there is a single or multiple agonist abnormality, she said. Reduced maximal aggregation with two or more agonists is associated with a bleeding disorder, and reduced MA with a single agonist (except with ristocetin and collagen) often represents a false-positive.
In an example of MA findings for two patients with multiple LTA abnormalities, the patient with Glanzmann’s shows absent aggregation response to all agonists except ristocetin (patient response is 47 with 1.25 mg/mL ristocetin; reference interval: >75 for adjusted PRP). “The response to ristocetin is reflective of agglutination without aggregation, and that’s why that response is also reduced,” Dr. Hayward said.
In the patient who has a thromboxane generation defect, in this case due to aspirin, there is the hallmark feature of a striking reduction in arachidonic acid response (patient response is 6 with 1.6 mM arachidonic acid; reference interval: >77 for adjusted PRP) with a normal response to thromboxane analogue (patient response is 94 to 1 μM thromboxane analogue; reference interval: >70 for adjusted PRP).
“When there’s a thromboxane generation defect,” Dr. Hayward said, “it is often associated with reduced maximal aggregation collagen. With epinephrine in this case, there is no secondary aggregation, and that is also part of the phenotype of an aspirin-like defect.”
In cases of Glanzmann’s, Bernard Soulier syndrome, or an aspirin-like defect, for example, the possibility of an inherited or an acquired defect should be raised at the time the test finding is interpreted, she said. And confirmatory testing should be considered if, in recognizing patterns that are potential false-positives, an abnormal pattern is seen with multiple agonists.
[dropcap]D[/dropcap]r. Hayward addressed what to do when evaluating aggregation findings for a patient who has a significant thrombocytopenia and the patient’s platelet-rich plasma sample has a low platelet count.
“We like to use a checklist of which agonist to test based on the sample platelet count,” she said. “Some agonists, such as epinephrine, become noninformative when the platelet count of the sample is quite low.”
“Now what about the control?” she asked. “We compare the data for such cases with reference intervals that we prepared using regression to estimate the 95 percent confidence intervals for maximal aggregation responses based on sample platelet counts. And so we take the plots from our published paper of what we’re expecting to see, and we plot our actual patient’s results on those graphs,” she said.
In the aggregation findings for one case, the patient had significantly impaired aggregation with collagen but also with ristocetin. “This suggested to me a particular diagnosis,” Dr. Hayward said. “This was a case where we had a mother and a son with a thrombocytopenic platelet disorder and bleeding symptoms.”
In this case, they did flow cytometry to further evaluate this person’s findings. The glycoprotein analysis showed that the patient had normal expression of the von Willebrand factor receptor but significantly reduced expression of the platelet fibrinogen receptor, αIIbβ3, also called glycoprotein IIb/IIIa, she said. Subsequent molecular studies confirmed that she had variant Glanzmann thrombasthenia, due to an activating mutation in this platelet glycoprotein receptor.
“This is an autosomal dominant type of platelet function disorder,” she added.
Dr. Hayward also addressed the importance of correlating what is seen in the laboratory with the clinical picture. “If your lab tests for dense granule deficiency and for aggregation defects, and the findings are abnormal but it’s not a clear-cut syndrome, what does that mean for the patient who has confirmed dense granule deficiency or aggregation defects?”
She and colleagues have done research to address this. She presented data from a recent study of a cohort with such nonsyndromic platelet function disorders that manifest with either confirmed dense granule deficiency or aggregation defects with multiple agonists, due to pathogenic RUNX1 mutation or unknown molecular causes. “The latter are much more common” (Brunet J, et al. Res Pract Thromb Haemost. 2020;4[5]:799–806).
These individuals have increased risk for developing different types of mucocutaneous bleeding problems, Dr. Hayward said. “For example, their risk of needing a blood transfusion for bleeding is 100 times higher than people in the general population. They are at higher risk for experiencing minor wound bleeding and wound healing problems. They have a 20-fold increased risk of bleeding with surgery and a 12-fold increased risk of bleeding with serious accidents or trauma.” These patients are also at increased risk for bleeding with a urinary tract infection and for gastrointestinal bleeding.
“When we did our study, we noticed there were differences between the women and the men with platelet function disorders,” Dr. Hayward said. “The women had more mucocutaneous bleeding, like bruising,” and additional bleeding with challenges such as childbirth, miscarriages, and menses.
“Their increased risk for bleeding with childbirth is 17-fold. Their risk for having prolonged menses is increased almost ninefold. Compared to men, women with these disorders are more likely to present with bruising, bruising for no reason, and bruising disproportionate to trauma, and this may be why when we’re seeing people with these disorders, there’s a referral bias to see many more women than men.”
Amy Carpenter Aquino is CAP TODAY senior editor.