• platelet function analyzer

We read with interest the critical comments of Drs Elalamy and Hatmi on our recent paper (Haubelt et al, 2005). They maintain that the higher upper limit of the references range for collagen/adrenaline (CEPI) closure times (CT) found in our study resulted from platelet dysfunction due to unreported drug intake and included thyroid hormones or dietary components which might impair platelet function as possible factors that might have influenced CEPI CT.

In light of this criticism, we re-examined our methods and data and have concluded that our CEPI-CT results were not influenced significantly by any of these factors. All individuals that reported a history of bleeding or intake of drugs influencing platelet function on a standardised questionnaire had been excluded. In addition, and in contrast to most previous studies establishing referenced ranges for Platelet Function Analyzer 100 (PFA-100TM) CT, we measured arachidonic acid (AA)-induced platelet aggregation (PA). This very sensitive assay for detecting aspirin responsiveness usually results in differences in light transmission clearly below 20%.

All individuals included in our study were under a normal Western diet and had AA-induced PA results markedly above the threshold value of 20%. Moreover, it is generally known that platelet function assay results vary widely, even among healthy individuals examined under comprehensively standardised conditions.

Drs Elalamy and Hatmi cited a paper (Mamiya et al, 1989) that suggests that thyroid hormone supplementation impairs platelet function. But there have been no reported observations of clinically relevant platelet dysfunction in euthyroid individuals under administration of thyroid hormones (Myrup et al, 1995; Masunaga et al, 1997). Apparently, the in vivo effects of orally administered thyroid gland hormones are not comparable with the in vitro effects on washed human platelets reported in the cited study.

Two of our 120 volunteers were indeed taking 100 μg/d l-thyroxine at the time of examination. But we were able to show that a higher CEPI-CT could be traced to a manufacturing fault in the respective lot of CEPI cartridges. Re-examination of 11 individuals with CEPI-CT exceeding 200 s using another batch of CEPI cartridges revealed significantly shorter CEPI-CT (medians, 253 s vs. 161 s), while collagen/ADP (CADP)-CT, bleeding time (BT), PA and von Willebrand factor-collagen binding activity (VWF:CBA) were not significantly different between the two measurements. Moreover, subjects with a CEPI-CT above the 90th percentile had significantly longer CADP-CT and significantly lower VWF:CBA levels than the remaining 108 subjects. BT and PA, however, did not differ significantly between individuals with CEPI-CT above or below the 90th percentile.

We concede that duplicate testing does not result in substantial improvement of accuracy. However, numerous differences in duplicates exceeding 10% or 15% are clearly due to manufacturing faults in the cartridges or the machine, resulting in variable reaction conditions. As a result, duplicate measurements are essential to detect outliers, which might potentially lead to misdiagnosis and inappropriate therapy.

Unlike Elalamy and Hatmi, we believe that the time of day when analyses should be performed has to be carefully standardised for every assay, whose results are subject to diurnal variation.

It is true that our females had significantly greater BMI values than males. However, only five of our 60 female and six of our male volunteers had BMI values above 30, and the blood group O/non-O ratios were not significantly different between groups (P = 0·069). It remains to be determined whether the slight but significant difference in CADP-CT but not in CEPI-CT, BT, PA and VWF:CBA is associated with greater BMI in women. The cited study (Böck et al, 1999) was merely observational, not a controlled trial.

We agree that the PFA-100 is more sensitive for detecting aspirin intake and von Willebrand factor deficiencies than the BT. However, the PFA-100TM system has several shortcomings, not least of which is a substantial number of false positive and false negative results. It also has low sensitivity for detecting mild non aspirin-induced platelet dysfunction (Wuillemin et al, 2002) or platelet dysfunction induced by ADP receptor antagonists (Golanski et al, 2004).


  1. Top of page
  2. References
  • Böck, M., de Haan, J., Beck, K-H., Gutensohn, K., Hertfelder, H-J., Karger, R., Heim, M.U., Beeser, H., Weber, D. & Kretschmer, V. (1999) Standardization of the PFA-100® platelet function test in 105 mmol/l buffered citrate: effect of gender, smoking, and oral contraceptives. British Journal of Haematology, 106, 898904.
  • Golanski, J, Pluta, J., Baraniak, J. & Datala, C. (2004) Limited usefulness of the PFA-100 for the monitoring of ADP receptor antagonists – in vitro experience. Clinical Chemical Laboratory Medicine, 42, 2529.
  • Haubelt, H., Anders, C., Vogt, A., Hoerdt, P., Seyfert, U.T. & Hellstern, P. (2005) Variables influencing Platelet Function Analyzer-100TM closure times in healthy individuals. British Journal of Haematology, 130, 759767.
  • Mamiya, S., Hagiwara, M., Inoue, S. & Hidaka, H. (1989) Thyroid hormones inhibit platelet function and myosin light chain kinase. Journal of Biological Chemistry, 264, 85758579.
  • Masunaga, R., Nagasaka, A., Nakai, A., Kotake, M., Sawai, Y., Oda, N., Mokuno, T., Shimazaki, K., Hayakawa, N., Kato, R., Hirano, E., Hagiwara, M. & Hidaka, H. (1997) Alteration of platelet aggregation in patients with thyroid disorders. Metabolism: Clinical and Experimental, 46, 11281131.
  • Myrup, B., Bregengard, C. & Faber, J. (1995) Primary haemostasis in thyroid disease. Journal of Internal Medicine, 238, 5963.
  • Wuillemin, W.A., Gasser, K.M., Zeerleder, S.S. & Lämmle, B. (2002) Evaluation of a Platelet Function Analyser (PFA-100® in patients with a bleeding tendency. Swiss Medical Weekly, 132, 443448.