November 2011

Editor:
Fredrick L. Kiechle, MD, PhD

Q. In our laboratory in the past, if we had a flagged automated platelet count, we performed a platelet estimate, and if the two matched, we reported the automated count. If, however, the estimate did not match, usually owing to giant platelets, we per-formed a manual platelet count using the Unopette (BD) product. Since the product is no longer avail-able, we would like to update our platelet estimation method to eliminate the manual platelet count completely. We have found great discrepancies in the literature in the various methods used. None seem to come close to the actual count. One reference said giant platelets should not be counted in the estimate because they are most likely impaired in their function. How can we establish an accurate platelet estimate for our laboratory? Does the presence of giant platelets alter how the estimate should be done? By what percentage should the estimated count agree with the automated count?

A. Katherine A. Galagan, MD, answered a related question on platelet estimates in the February 2010 issue of CAP TODAY. In that question, the reader asked if platelet estimate is an acceptable method when the platelet count is low and how the estimated platelet count should be reported.¹

Platelet counting can be challenging in some patients because of characteristics that overlap with other cellular material, such as schistocytes and leukocyte cytoplasmic fragments, the presence of cryoglobulins, and the inherent ability of platelets to activate and clump. In addition, as mentioned, giant platelets may not be counted as platelets by automated analyzers because of their size. Though automated hematology analyzers usually provide accurate platelet counts with generally good precision, in some circumstances interference with the automated count can occur, requiring a manual method of platelet analysis. Platelet estimate from the blood smear is an acceptable method for counting platelets in many situations, including those in which similar sized particles interfere with the automated count. A phase platelet count can also be used in similar circumstances, though white cell cytoplasmic fragments can be difficult to distinguish from platelets by this method. As mentioned, the Unopette product is no longer available for this purpose. An available alternative for phase platelet analysis is the Thrombo-TIC product (Bioanalytic GmbH, Germany). Giant platelets should be included in the platelet count by these methods. Unfortunately, platelet clumping precludes accurate manual methods as well because of the very uneven distribution of platelets. If repeat blood draw with citrate anticoagulation does not prevent platelet clumping, a bedside fingerstick for phase platelet count can be useful.

As Dr. Galagan said in her response, platelet estimate from the blood smear is generally performed by one of two methods. The first method involves counting platelets in 10 or more fields at 100× magnification, including thin and thick areas if the distribution is uneven, and then multiplying by a “field factor” to account for the size of the microscopic field. Determining the size of the microscopic field of view at 100× magnification is important in determining the number of fields to count and the appropriate field factor. Variability by this method can occur due to the patient’s degree of anemia, with the tendency to assess too deep (too far from the feathered edge) on the smear in anemic patients. Accuracy can be improved with consistent smear preparation and assessment of similar regions of the smear in all patients. Generally, these platelet estimates will be within about 15 percent of automated counts.² However, at very low levels, the coefficient of variation for the platelet estimate is high and may result in increased discordance with an analyzer result. For platelet estimate, one study observed a CV of 50 percent at 27 × 109/L, 20 to 23 percent at 65 to 100 × 109/L, and nine to 15 percent at 163 to 590 × 109/L.³ A second method for manual platelet counting involves counting platelets in an area containing 1,000 red cells, with the platelet count (× 109/L) being the number of platelets counted multiplied by the red blood cell count.⁴ This method, by design, accounts for the patient’s degree of anemia, though variability can also be seen if platelet distribution is uneven on the blood smear. Based on the imprecision of the platelet estimate on the blood smear, especially at the low levels important for clinical decisionmaking, many laboratories choose to report “semiquantitative” results with numerical ranges rather than an exact number.

References

Joan E. Etzell, MD
Director, Clinical Hematology Laboratory
University of California
San Francisco

Vice Chair, CAP Hematology/
Clinical Microscopy Resource Committee

Q. What is the future of LC-MS/MS in the routine clinical testing laboratory?

A. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is one of many ways to achieve separation (LC) and detection of ions (MS) based on mass and charge. MS technology characteristically represents an ionic footprint of a large molecule and now, mirroring what molecular DNA-based technology has brought to the clinical laboratory, is used routinely in the clinical laboratory to quantify a number of molecules and analytes from complex serum and plasma matrices. A number of clinical laboratories use LC-MS/MS for quantification of therapeutic drugs, drugs of abuse, and steroid hormones and to screen newborns for inborn errors of metabolism.¹ In addition, other recent literature suggests that MS technology, like matrix-assisted laser desorption/ionization–time of flight mass spectrometry (MALDI-TOF MS), has applications in the clinical laboratory for direct bacterial identification and classification.^2,3

MS has been a mainstay technology in the environmental, pharmaceutical, and life/biomedical science markets, and it has been commonplace in such industries and in academic research institutions for more than a century. Only in the past 25 to 35 years have we seen MS technology break into the clinical laboratory and, as of late, with increased momentum. Key components to bringing MS into the clinical laboratory are explained in a recent Clinical Laboratory News article.¹ While MS technology is expensive, it is not cost-prohibitive if you do your homework, use your expertise and skills, work collaboratively with your vendor, and take the time to implement, validate, and become comfortable with the technology. In theory, validating and implementing an MS assay is no different from any other test in the clinical laboratory, and there are approved guidelines for clinical MS method development and validation. However, a recent review of the literature warns laboratorians about common pitfalls, potential inaccuracies, and sources of error related to using LC-MS/MS technology.^4-6

The success of MS as a technology used routinely in the clinical laboratory depends on a number of things. Vendors and manufacturers need to supply the clinical laboratory with fully validated, FDA-approved LC-MS/MS methods, support the laboratory by providing skills training and service, improve MS design so instruments can be compact, develop software that simplifies operation and instrument control, and provide a means to interface and integrate the flow of information from the MS instrument to clinical information systems. The clinical laboratory community must continue to demand that MS vendors develop systems, product lines, and software that suit the clinical laboratory environment. Until then, MS will be anything but routine.

References

Deanna Franke, PhD, DABCC
Clinical Scientist
Core Laboratory and Toxicology
Pathology Consultants of South Broward
Memorial Healthcare System
Hollywood, Fla.