- Email: [email protected]
Hemostasis in neonates and children: Pitfalls and dilemmas
Blood Reviews 24 (2010) 63–68
Contents lists available at ScienceDirect
Blood Reviews journal homepage: www.elsevier.com/locate/blre
REVIEW
Hemostasis in neonates and children: Pitfalls and dilemmas Paul Monagle a,*, Vera Ignjatovic b,1, Helen Savoia c,2 a
Department of Clinical Haematology, Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Flemington Rd. Parkville, Victoria 3052, Australia Haematology Research, Murdoch Children’s Research Institute, Royal Children’s Hospital, Flemington Rd. Parkville, Victoria 3052, Australia c Laboratory Services, Royal Children’s Hospital, Flemington Rd. Parkville, Victoria 3052, Australia b
a r t i c l e Keywords: Hemostasis Neonate Child Development
i n f o
s u m m a r y Developmental Hemostasis refers to the age-related changes in the coagulation system that are most marked during neonatal life and childhood. An understanding of these changes is crucial to the accurate diagnosis of hemostatic abnormalities in neonates and children. This paper explains the current understanding of developmental hemostasis and describes the common pitfalls observed in clinical practice through failure to implement the principles into routine diagnostic work. Finally, there is a brief discussion as to a potential physiological rationale for developmental hemostasis and the implications of this for hemostatic interventions in neonates and children. There remains a need for further study to improve our understanding of the implications of developmental hemostasis in normal growth and development. Ó 2009 Elsevier Ltd. All rights reserved.
Introduction
Developmental hemostasis
Hemostasis is a complex homeostatic system which is critical to maintain healthy life. Disturbances of the hemostatic system account for a large proportion of the non-infectious morbidity and mortality in westernized societies, and are integral to many disease processes including cancer. At a simple level, the critical components of hemostasis are still best described by Virchow’s triad of blood vessel wall, blood composition and blood flow. In considering the non-flow aspects of this equation, hemostasis can be considered as an interaction between the blood vessel wall, coagulation proteins within the plasma, and blood cellular components, predominantly platelets. The hemostatic system interacts with other physiological systems such as wound repair, inflammation and angiogenesis, and the extent and nature of these interactions are still being defined. For the purposes of this paper, the discussions of the hemostatic system will be confined to the coagulation proteins within the plasma. While recognizing the importance of the blood vessel wall and blood cellular interactions, the role of these components are in many ways far less understood at this stage. However, given the nature of clinical laboratory testing, they may also present fewer diagnostic dilemmas for practicing clinicians.
Hemostasis is a dynamic, evolving process that is age-dependent and begins in utero. The evolution of the hemostatic system continues throughout life, however the changes are most marked during childhood, and hence of most clinical relevance during this time. Although evolving, the hemostatic system in healthy fetuses, infants and children must be considered physiologic. The term ‘‘Developmental Hemostasis” was coined in the late 1980s by Dr. Maureen Andrew to describe this phenomenon. Dr. Andrew demonstrated, in a single large cohort study of Canadian children that the concentrations of the majority of coagulation proteins, as measured by functional assays, varied significantly with age. The landmark papers, published in Blood in 1987, 1988 and 1992, not only changed the approach to research related to the coagulation system in children, but also the clinical interpretation of coagulation assays used by many hemostasis laboratories around the world since that time.1–3 However, almost three decades later, there remain many gaps in our understanding of the mechanisms, implications of, and reasons for these observed age-related changes. The evaluation of newborns and children for hemorrhagic or thrombotic complications presents many potential pitfalls and dilemmas that are not encountered in adults. This paper aims to describe the current state of knowledge with respect to developmental hemostasis and then discuss the common pitfalls and strategies to avoid these problems. An understanding of developmental hemostasis in the broadest sense optimizes the prevention, diagnosis, and treatment of hemostatic problems during childhood and undoubtedly provides new insights into the pathophysiology of hemorrhagic and thrombotic complications for all ages.
* Corresponding author. Tel.: +61 3 9345 5914; fax: +61 3 9349 1819. E-mail addresses: [email protected] (P. Monagle), [email protected] (V. Ignjatovic), [email protected] (H. Savoia). 1 Tel.: +61 3 9936 6520; fax: +61 3 8341 6212. 2 Tel.: +61 3 9345 5914; fax: +61 3 9349 1819. 0268-960X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.blre.2009.12.001
64
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
Age-related changes in the coagulation system The fundamental principle underlying developmental hemostasis is that the functional levels of coagulation proteins change in a predictable way with changes in age.4 While the absolute values of these changes are reagent and analyzer dependent, the trends observed are consistent across a number of studies.5 These changes in functional protein levels lead to corresponding changes in global tests of coagulation such as the Activated Partial Thromboplastin Time (APTT). Other global measures of hemostasis may be more or less sensitive to age-related changes. For example, there is little difference in normal thromboelastography (TEG) with age.6 Studies to date have predominantly involved functional assays of the coagulation proteins due to the fact that these assays are commonly used in clinical practice. Whether these changes represent true changes in plasma protein concentrations, or functional changes associated with post-translational protein modifications, or the presence or absence of other as yet unidentified co-factors remains unknown. Previous studies have confirmed that post-translational modifications in coagulation proteins do occur with age. For example, fibrinogen has previously been demonstrated to exist in a ‘‘fetal” form, in cord blood of term infants.7 This ‘‘fetal” fibrinogen was shown to have increased sialic acid content compared to adult fibrinogen, a direct result of post-translational modification. Moreover, in 1970 Witt et al. reported a twofold increase in the phosphorus content of foetal fibrinogen in comparison to that measured in adult fibrinogen.8 These findings were corroborated by a subsequent report of 3–4-fold increase in phosphorus content in fetal fibrinogen compared to adult fibrinogen.9 In addition, specific thrombin clotting times were prolonged in newborns suggesting differences in polymerisation of fibrin from ‘‘fetal” fibrinogen,10 an observation that has lead to claims that fibrinogen in infants is ‘‘dysfunctional”.11 Observations that an increase in sialic acid content of fibrinogen is associated with a decreased rate of fibrin polymerisation while the removal of sialic acid residues leads to increase in polymerisation12 provide a possible explanation for the differences in thrombin clotting times. Definite evidence of the importance of differences in sialic acid content of fibrinogen is provided by observations that sialic residues of fibrinogen directly bind Ca2+.13 This binding leads to a decrease in the intermolecular repulsion between the fibrinogen chains and via that mechanism facilitates fibrin polymerisation.13 Whether such changes exist in older children, and the extent to which similar changes (phosphorylation, glycosylation etc) exist in other hemostatic proteins remains to be determined. The impact of these biochemical changes on routine clinical hemostatic assays, and therefore developmental hemostasis as reflected in routine clinical diagnostic assays is unclear. Table 1 lists major studies that have described age-related changes in coagulation proteins and routine coagulation assays. This paper will not describe all the reported changes in detail, but refers the reader to the papers cited in Table 1. Despite the changes in individual protein levels and in global tests of coagulation, the hemostatic system in neonates and children does not seem disadvantageous compared to the ‘‘normal” coagulation system as measured in adults. There is no data to support either an increased bleeding or thrombotic risk during infancy and childhood for any given stimulus and on the contrary, one could argue that the hemostatic system in neonates and children is protective against bleeding and thrombotic complications compared to adults. This is despite the fact that when considering individual proteins, many proteins exist at levels during stages of infancy that would be associated with disease in adults. For example, normal Antithrombin (AT) levels during the first three months
of life are lower than that observed in many adults with heterozygous AT deficiency and recurrent thrombosis.7 Similarly levels of Factor IX may be below 30% of adult levels in early life and yet these children have no evidence of a bleeding phenotype. Prolongation of global tests of coagulation, such as the APTT, are also not associated with any increase in bleeding, despite the APTT being considerably longer compared to that observed in many adults with clinically relevant VWD.7 The absolute levels of hemostatic proteins as measured by the currently available functional assays are clearly not the key factor in determining clinical phenotype, and there remains much research to be performed to better understand the relationships between levels of hemostatic proteins and functional outcomes of the hemostatic system.
Pitfalls and dilemmas interpreting hemostatic testing in neonates and children Sample acquisition Pre-analytical variables are known to be important in coagulation testing for any age-group. However, nowhere is this more problematic than in neonates and children. Obtaining sufficient blood via clean venepuncture can be extremely challenging in infants. Frequently bloods are obtained via central venous or arterial access devices, where heparin contamination or sample activation are common place. Blood may be collected into a pre-heparinised syringe prior to being transferred to a citrate tube. Further, many neonates will have a relative polycythaemia, meaning that the desired citrate-to-blood ratio (9:1) may not have been achieved, and collection into an adjusted tube is required. Alternatively, collection tubes are frequently under-filled, which is likely not significant for a chemistry panel or a full blood examination, but will ultimately produce inaccurate coagulation results if the citrateto-blood ratio has once again been altered.
Defining reference ranges A number of studies have confirmed that age-related reference ranges for coagulation assays in the healthy population are analyzer and reagent dependent (see Table 1). Yet, very few clinical laboratories have adequately determined age-related reference ranges specific to their analyzer–reagent combination. This deficiency is most likely due to the cost and ethical difficulties in collecting blood from otherwise healthy children to enable reference range determination. Thus, many laboratories compare their pediatric results to either their adult reference range or to published pediatric reference ranges. Neither of these options is satisfactory in terms of the highest standard of clinical care of the patients. Take for example a 4 year-old child who has an APTT performed, using PTT-A reagent (Diagnostica Stago, France) on an STA- compact analyzer (Diagnostica Stago, France). Ninety-five percent confidence interval for a reference range in this age-group is 33–44 s (rounded to whole seconds). If in contrast, the reagent being used is CK-PREST (Diagnostica Stago, France), even with the same analyzer, the reference range encompassing 95% of the population is reported to be 30–35 s.7 Thus if one was referencing to a published reference range based on CK-PREST analysis, but using PTT-A as your reagent, then over 30% of children (assuming normal distribution) in this age-group would be labeled as having prolonged APTT when in fact they were perfectly healthy. Given that the majority of APTT reagents currently in clinical use have no published age-related references ranges, such errors are likely to be frequent. As each new reagent is released onto the market, unless the appropri-
65
P. Monagle et al. / Blood Reviews 24 (2010) 63–68 Table 1 Studies reporting age-related differences in coagulation assays or proteins during childhood. Author
Year
Assays/proteins reported
Age-groups
N
Perlman
1975
PT, TT, APTT, fibrinogen, FDP, platelet count, haematocrit, FV, FVIII, plasminogen, hemoglobin
Healthy infants Small-for-dates infants Post-mature infants
n = 35 n = 26 n = 30
Beverley, D.W. et al.
1984
APTT, FII-VII-X, fibrinogen, a2-antiplasmin, platelet count, MPV, megathrombocyte index, plasminogen
Cord blood Newborns (48 h)
n = 80
Andrew, M. et al.
1987
PT, APTT, TCT, fibrinogen, FII, FV, FVII, FVIII, vWF, FIX, FX, FXI, FXII, PK, HMW-K, FXIIIa, FXIIIb, plasminogen, antithrombin, a2-M, a2-AP, C1E-INH, a1-AT, HCII, protein C, protein S
Day 1 newborn Day 5 newborn Day 30 newborn Day 90 newborn Day 180 newborn Adult
28–75 samples per agegroup
Andrew, M. et al.
1988
PT, APTT, TCT, fibrinogen, FII, FV, FVII, FVIII, vWF, FIX, FX, FXI, FXII, PK, HMW-K, FXIIIa, FXIIIb, plasminogen, antithrombin, a2-M, a2-AP, C1E-INH, a1-AT, HCII, protein C, protein S
Premature newborns (30–36 weeks gestation) Day 1 Day 5 Day 30 Day 90 Day 180
23–67 samples per agegroup
Andrew, M. et al.
1992
PT/INR, APTT, bleeding time, fibrinogen, FII, FV, FVII, FVIII, vWF, FIX, FX, FXI, FXII, PK, HMW-K, FXIIIa, FXIIIs, plasminogen, TPA, PAI, antithrombin, a2-M, a2-AP, C1E-INH, a1-AT, HCII, protein C, protein S (total and free)
1–5 years 6–10 years 11–16 years Adults
20–50 samples per agegroup
Reverdiau-Moalic, P. et al.
1996
PT/INR, APTT TCT, FI, FII, FVII, FVII, FIX, FX, FV, FVIII, FXI, FXII, PK, HMWK, AT, HCII, TFPI, protein C (Ag, Act), protein S (free and total), C4b-BP
Fetuses 19–23 24–29 30–38 weeks gestation Newborns (immediately after delivery) Adults
Cargo, M.D. et al.
2002
PFA100 Hb platelet count
Neonates Children Adults
n = 17 n = 57 n = 31
Salonvaara, M. et al.
2003
FII, FV, FVII, FX, APTT, PT/INR, platelet count
Premature infants 24–27 weeks 28–20 weeks 31–33 weeks 34–36 weeks
n = 21 n = 25 n = 34 n = 45
Flanders, M. et al.
2004
PT, APTT, FVIII, FIX, FXI, antithrombin, RCF, vWF, protein C, protein S
7–9 years 10–11 years 12–13 years 14–15 years 16–17 years Adults
124 per age-group
Monagle, P. et al.
2006
APTT (4 reagents), PT/INR, fibrinogen, TCT, FII, FV, FVII, FVIII, FIX, FX, FXI, FXII, antithrombin, protein C, protein S, D-dimers, TFPI (free and total), endogenous thrombin potential
Day 1 Day 3
Minimum 20 samples per age-group
n = 20 n = 22 n = 22 n = 60 n = 40
<1 years 1–5 years 6–10 years 11–16 years Adults Chan, K.L. et al.
2007
Thromboelastography (TEG) [R. K. a, MA, LY30]
<1 years 1–5 years 6–10 years 11–16 years Adults
n = 24 n = 24 n = 26 n = 26 n = 25
Sosothikul, D. et al.
2007
PT, APTT, fibrinogen, TAT, PC:Ac, TF, FVIIa, sTM, vWF (Ag & RCo), D-dimer, tPA, PAI-1, TAFI
1–5 years 6–10 years 11–18 years Adults
n = 19 n = 26 n = 25 n = 26
Mitsiakos, G. et al.
2008
INR, PT, APTT, fibrinogen, FII, FV, FVII, FVIII, FIX, FX, FXI, FXII, antithrombin, protein C, protein S, APCr, tPA, PAI-1, VWF
Small for growth newborns Appropriate for growth newborns
n = 90 n = 98
Newall, F. et al.
2008
PF4 and Vitronectin
<1 years 1–5 years 6–10 years 11–16 years Adults
15 per age-group
(continued on next page)
66
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
Table 1 (continued) Author
Year
Assays/proteins reported
Age-groups
N
Ries, M. et al.
1997
TAT, F1+2, PAP, D-dimer
1–6 years 7–12 years 13–18 years Adults
20 per age-group
Boos, J. et al.
1989
PIVKA II, FVII, FII, FII:Ag
Day 1, 2, 3 neonates
n = 57 total
ate normative data is generated and published, then children will continue to be classified as either normal or abnormal erroneously. Similar outcome is observed, if an adult based reference range is used. For example, the adult reference range for an APTT using PTTA reagent (Diagnostica Stago, France) and an STA-Compact analyzer is 27–38 s (95% confidence intervals rounded to whole seconds). Once again approximately 30% of children aged 1–5 years would be classified as abnormal if they were compared to this adult reference range, even if their samples were analyzed using the same reagent and analyzer combination. The clinical implications of this type of error are enormous. If one considers the most likely clinical indications for performing an APTT assay in non-hospitalized children, then the true cost of this misclassification starts to become apparent. Children commonly have coagulation studies ordered because they are preoperative and the surgeon is concerned about the past history (‘‘my child bruises easily, or has blood noses”), or because of a positive family history. Erroneous interpretation of an APTT as prolonged, due to the use of an inappropriate reference range, leads to cancellation of surgery, multiple investigations, referral to specialist hematologists and often over-treatment of the child during the procedure. Similarly, children often have coagulation studies because of a positive family history of a bleeding disorder. Misclassification of the child due to inappropriate reference ranges can lead to a child being labeled as having a bleeding or thrombotic disorder when in reality they do not. This can have significant lifestyle and other implications for the child and their family which can be problematic for many years to come. Reversing an already made ‘‘diagnosis” is never easy. Finally, children thought to be at risk from physical or mental abuse will often have coagulation studies as part of their workup if they present with unusual bruising or injuries. Trying to explain in a court of law that you do not believe a child has a bleeding disorder, and that the bruises have truly been a result of child abuse, when the APTT or other coagulation assay results lie outside the reference range quoted by your laboratory is very difficult and may contribute to a miscarriage of justice and further danger to the child. In summary, the lack of use of appropriate reference ranges remains one of the most common pitfalls in the interpretation of pediatric coagulation assays. This practice contributes to inappropriate investigation of children, as well as erroneous diagnosis and management of often healthy children. All laboratories which report coagulation studies in neonates and children must ensure that they are using appropriate age-related reference ranges that are specific for their analyzer–reagent combination. Diagnosing disease Of course, having a value for a hemostatic test, which is outside of the defined 95% confidence intervals for age does not necessarily define disease. Herein exists another dilemma for assessment of hemostasis in neonates and children. For example, if the lower limit of the reference range for plasma levels of factor VIII in a child is 59% based on a population study, and a patient of appropriate age presents who, when using the same analyzer and reagent, has a plasma level for factor VIII of 50%, then it is unlikely that the pa-
tient has even mild hemophilia. This is because by definition, we define hemophilia by arbitrary plasma levels (<30% for mild hemophilia), or at least plasma levels that are thought to be associated with increased risk of bleeding. In the absence of genetic studies definition of heterozygote states remains difficult. The dilemma is that for the laboratory tests to be clinically useful, we need to understand the level at which results define clinically relevant disease probably more so than the levels at which the results are outside the 95% confidence limits of the healthy population. To date, there have been no specific studies in children which address this issue. Correlation of phenotype and genotype, as well as longitudinal studies mapping the hemostatic protein changes in a cohort of patients in conjunction with their clinical outcomes would be most helpful. The issue of results being outside the reference range versus being indicative of clinical disease is most problematic when considering the procoagulant proteins. One must always consider that a positive diagnosis of a hemostatic disorder should include the presence of a family history, a positive clinical phenotype, and reproducible abnormal laboratory results. The problem is that the clinical phenotype is much more difficult to define given the infrequency of thrombotic events; the family history is often equivocal in terms of the role of clinical precipitants of thrombosis, and yet the labeling of children as having a thrombophilic condition can have significant impact. Frequently clinicians are asked to advise on the need for thrombopropylaxis related to clinical interventions, oral contraceptive pill use, or prolonged travel in children in whom the presence or absence of a true thrombophilic condition seems almost impossible to determine. New diagnostic assays The quest for improved clinically predictive value from diagnostic assays is never ending. Over recent years, thrombin generation assays of various types have been promoted as the ideal coagulation assays. In theory, based on our current understanding of the hemostatic system, this seems logical, as thrombin must be viewed as the key enzyme within the hemostatic system and hence understanding the level of thrombin activity should be most helpful in determining whether the hemostatic system is primed for clotting or susceptible to bleeding. The Endogenous Thrombin Potential (ETP) provides an in vitro measure of the overall ability to generate thrombin, which in turn converts fibrinogen to fibrin. As stated, the ETP therefore provides perhaps the best assessment of global hemostasis.14 Currently available chromogenic and fluorogenic substrates available for ETP measurement are cleaved by both free (active) and alpha-2-macroglobulin-bound (inactive) thrombin, leading to an overestimation of thrombin generation. The currently available commercial methods, such as automated fluorogenic assays, account for this using a mathematical algorithm which assumes the likely contribution of alpha-2-macroglobulin to the ETP.15–17 Studies of developmental hemostasis have documented physiological variation in alpha-2-macroglobulin levels with age, where increased levels are observed throughout infancy and childhood compared to adults.1–3,5 The relative contribution of alpha-2macroglobulin bound to and inhibiting thrombin is significant and has been reported to vary from 16% to 64% across the pediatric age-
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
range, while being approximately 7% in adults.18 All methods of thrombin generation measurement must therefore account for this significant, inactive thrombin complex. In particular, the variation with age must be accounted for to prevent over-estimation in the younger populations, where the alpha-2-macroglobulin component plays a significant role. Automated methods that do not directly measure the alpha-2-macroglobulin bound component, but instead rely on algorithms derived from adult populations, are unlikely to account for the significant developmental variation in alpha-2-macroglobulin across the age. Thus, methods which depend on preset algorithms are unlikely to provide accurate results in children of all ages.18 This example demonstrates that the introduction of all new assays into clinical use in children should be based on specific clinical studies in children, confirming the clinical predictive value of the assays, or at the very least should be based on a thorough understanding of the developmental changes in the hemostatic system. Use of clinical algorithms As discussed previously, the use of diagnostic assays, the results of which depend on mathematical calculations based on assumptions formulated in adult testing, are unlikely to be accurate in children. The same may be said of clinical decision making algorithms. Risk stratification for treatment or prophylaxis often depends on clinical algorithms which involve the results of laboratory assays. For example, algorithms which utilize the D-dimer in determining the likelihood of pulmonary embolism,19 Such algorithms have rarely been clinically tested in children, and the only studies addressing this issue have reported that the algorithms are not valid in children.20 The reasons for the lack of validity may include the age-related variation in the normal hemostatic system, uncertainties about the ability to truly diagnose asymptomatic hemostatic disease and probably the pathophysiological differences in bleeding and thrombosis development. However, the evidence to date would suggest that algorithms used in adults are unlikely to be valid in children without modifications, that need to be specifically tested in the pediatric population. This poses a constant dilemma for clinicians, who in the absence of specific pediatric data, often look to adult practice for guidance, and yet the basic principles on which the guidance is based is likely flawed. There remains an ongoing need for pediatric-specific studies to develop appropriate clinical algorithms. Monitoring anticoagulation in children Developmental hemostasis likely impacts on anticoagulant therapy in numerous ways, much of which is beyond the scope of this review. However, an understanding of the impact of developmental hemostasis on laboratory monitoring of anticoagulation is important. There have been no studies to determine target therapeutic ranges for any anticoagulant drug in children based on the gold standard of clinical outcome. The only studies performed to date have reported the doses required in children to achieve the therapeutic ranges determined for given indications in adults.21– 25 The likelihood that these therapeutic ranges are ideal in children seems small. First, given that the reference ranges for coagulation assays in healthy children are different to those in healthy adults, then, for example in the case of APTT monitoring of unfractionated heparin, the incremental increase in APTT to achieve the therapeutic range is clearly different.26 Second, the impact on thrombin generation for identical doses of anticoagulants in children versus adults has been shown to be different,27 suggesting that the target therapeutic ranges will likewise be different. Third, and perhaps most important, adult target therapeutic ranges for anticoagulation are based on clinical studies confirming that the target range
67
provides the optimal therapeutic benefit for the least risk (of bleeding).28,29 The different biology of thrombosis (related in part to developmental hemostasis) and the different risk of bleeding (related to the clinical state of the children) make it unreasonable to assume the target ranges will be identical for such vastly different patient populations. In addition, there is considerable data to show that there is a lack correlation of related monitoring assays, for example APTT versus anti-factor Xa assays for monitoring unfractionated heparin in children compared to adults and that the ability to distinguish ‘‘therapeutic” from ‘‘non-therapeutic” is reduced.26,30,31 Finally, the inter-assay variability seems to be exaggerated in children, and this can have significant clinical implications.32
The role of developmental hemostasis in normal physiology Perhaps the most intriguing aspect of developmental hemostasis is trying to understand the rationale for such marked age-related changes. The neonatal hemostatic system is arguably more protective against stimuli that might cause bleeding or thrombosis in adults. Hence, one could view the neonatal hemostatic system as the ideal system, with the effects of aging leading to deterioration in the system from an early age in healthy individuals. An alternative view is that the hemostatic system is so dramatically different in neonates and children for reasons unrelated to hemostasis. There is increasing evidence that the proteins involved in the hemostatic system have multiple functions in multiple physiological systems within the body, such as angiogenesis, inflammation and wound repair. One theory is that it is these systems that drive the developmental changes in the hemostatic system, which are therefore seen as necessary compensations to allow normal hemostatic function. Evidence for this theory is mounting, and one example of this is found in studies related to Antithrombin (AT). Recently, AT has been shown to have potent anti-angiogenic properties as well as that of being an important anticoagulant.33 AT has been shown to down regulate several pro-angiogenic genes and upregulate a number of anti-angiogenic genes. The anti-angiogenic forms of AT are known to include heparin binding sites, and heparin is shown to potentiate this effect.33 This is one of the postulated mechanisms in the positive effect of heparinoids on cancer survival which is independent of the antithrombotic effect.34–36 AT levels are naturally reduced in newborns to less than 50% of the levels observed in adults and then increase to approach adult levels by approximately six months of age.5 Early evidence suggests that there is a difference in the balance of isoforms of AT in newborns compared to adults. One postulated reason for the decreased levels of AT (and the altered balance of isoforms) observed in neonates is related to the role of this protein in angiogenesis.37 Fetal and early neonatal life is a time of prolific angiogenesis, much more so than any later stage of life, and given the known antiangiogenic properties of AT, then low levels of this protein are likely beneficial for healthy development. Hence, AT replacement therapy during neonatal life may well be deleterious by altering the normal balance of angiogenesis. In the only published randomized trial of AT replacement therapy in newborn infants (as a treatment for lung disease of prematurity), there were seven (11.5%) deaths in the AT-treated group and three (4.9%) deaths in the placebo (no treatment) group.38 A similar trial (reported only in abstract format) also reported a similar trend in the setting of a clinical trial. However, the mechanism for this finding was not able to be formally established.38 This is clearly an area that requires more research. The broader implications of manipulating the hemostatic system in children and infants is vital if we are to avoid non-hematological complications in these patients.
68
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
Conclusion In conclusion, developmental hemostasis is an important concept that has specific implications in the diagnosis of hemostatic disorders in children. Over-diagnosis and missed-diagnosis are both common when age-appropriate, analyzer and reagent specific reference ranges are not used. Further, as new assays are developed, or clinical algorithms utilized without a clear understanding of the impact of developmental hemostasis on their methodology and underlying principles, then the clinical utility of such assays/ algorithms in children will be uncertain. Finally, the rationale for the observed age-related changes in the hemostatic system remains unknown, and yet this would seem crucial to the determination of risk-benefit ratios for therapeutic manipulations of the hemostatic system in neonates and children. There remains an urgent need for more research into this crucial aspect of normal human development. Practice points Sample integrity is a major problem in pediatric coagulation studies. Attention to detail and use of repeat samples are important to avoid erroneous results. Age appropriate (analyzer and reagent specific) reference ranges are critical to the accurate diagnosis and management of coagulation disorders in children. Defining hemostatic disease requires more than a laboratory parameter outside of a population based reference range. Clinical algorithms used in adult practice need to be specifically tested in pediatric settings before they can be applied to neonates and children.
Research agenda The physiological causes responsible for the observed agerelated changes in coagulation proteins remain to be elucidated and are of tremendous biological significance. The development of pediatric-specific algorithms to guide clinical management requires specific clinical outcome studies in children. New assays need to be validated in children, in light of our understanding of developmental hemostasis.
Conflict of interest Paul Monagle and Vera Ignjatovic have received research funding from Diagnostico Stago, Roche, and Bayer. References 1. Andrew M, Paes B, Milner R, et al. Development of the human coagulation system in the healthy premature infant. Blood 1988;72:1651–7. 2. Andrew M, Paes B, Milner R, et al. Development of the human coagulation system in the full-term infant. Blood 1987;70:165–72. 3. Andrew M, Vegh P, Johnston M, Bowker J, Ofosu F, Mitchell L. Maturation of the hemostatic system during childhood. Blood 1992;80:1998–2005. 4. Andrew M, Paes B, Johnston M. Development of the hemostatic system in the neonate and young infant. Am J Pediatr Hematol Oncol 1990;12:95–104. 5. Monagle P, Barnes C, Ignjatovic V, et al. Developmental haemostasis. Impact for clinical haemostasis laboratories. Thromb Haemost 2006;95:362–72. 6. Chan KL, Summerhayes RG, Ignjatovic V, Horton SB, Monagle PT. Reference values for kaolin-activated thromboelastography in healthy children. Anesth Analg 2007;105:1610–3. [table of contents]. 7. Monagle P, Hagstrom J. Developmental haemostasis. In: Polin, editor. Fetal and neonatal physiology. 3rd ed. Elsevier: St. Louis; 2003. p. 1435–47. 8. Witt I, Muller H. Phosphorus and hexose content of human foetal fibrinogen. Biochim Biophys Acta 1970;221:402–4.
9. Hamulyak K, Nieuwenhuizen W, Devilee PP, Hemker HC. Reevaluation of some properties of fibrinogen, purified from cord blood of normal newborns. Thromb Res 1983;32:301–10. 10. Witt I, Muller H, Kunzer W. Evidence for the existence of foetal fibrinogen. Thromb Diath Haemorrh 1969;22:101–9. 11. Miller BE, Tosone SR, Guzzetta NA, Miller JLKK, Brosius KK. Fibrinogen in children undergoing cardiac surgery: is it effective? Anesth Analg 2004;99:1341–6. 12. Martinez J, MacDonald KA, Palascak JE. The role of sialic acid in the dysfibrinogenemia associated with liver disease: distribution of sialic acid on the constituent chains. Blood 1983;61:1196–202. 13. Dang CV, Shin CK, Bell WR, Nagaswami C, Weisel JW. Fibrinogen sialic acid residues are low affinity calcium binding sites that influence fibrin assembly. J Biol Chem 1989;264:15104–8. 14. Al Dieri R, Peyvandi F, Santagostino E, et al. The thrombogram in rare inherited coagulation disorders: its relation to clinical bleeding. Throm Haemost 2002;88:576–82. 15. Hemker HC, Beguin S. Thrombin generation in plasma: its assessment via the endogenous thrombin potential. Thromb Haemost 1995;74:134–8. 16. Hemker HC, Giesen P, Al Dieri R, Regnault V, de Smedt E, Wagenvoord R. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol Haemost Thromb 2003;33:4–15. 17. Hemker HC, Wielders S, Kessels H, Beguin S. Continuous registration of thrombin generation in plasma, its use for the determination of the thrombin potential. Thromb Haemost 1993;70:617–24. 18. Ignjatovic V, Greenway A, Summerhayes R, Monagle P. Thrombin generation: the functional role of alpha-2-macroglobulin and influence of developmental haemostasis. Br J Haematol 2007;138:366–8. 19. Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med 2001;135:98–107. 20. Biss TT, Brandao LR, Kahr WH, Chan AK, Williams S. Clinical probability score and D-dimer estimation lack utility in the diagnosis of childhood pulmonary embolism. J Thromb Haemost 2009;7:1633–8. 21. Bauman ME, Belletrutti MJ, Bajzar L, et al. Evaluation of enoxaparin dosing requirements in infants and children. Better dosing to achieve therapeutic levels. Thromb Haemost 2009;101:86–92. 22. Dix D, Andrew M, Marzinotto V, et al. The use of low molecular weight heparin in pediatric patients: a prospective cohort study. J Pediatr 2000;136:439–45. 23. Kuhle S, Massicotte P, Dinyari M, et al. Dose-finding and pharmacokinetics of therapeutic doses of tinzaparin in pediatric patients with thromboembolic events. Thromb Haemost 2005;94:1164–71. 24. Massicotte P, Adams M, Marzinotto V, Brooker LA, Andrew M. Low-molecularweight heparin in pediatric patients with thrombotic disease: a dose finding study. J Pediatr 1996;128:313–8. 25. Massicotte P, Julian JA, Gent M, et al. An open-label randomized controlled trial of low molecular weight heparin for the prevention of central venous linerelated thrombotic complications in children: the PROTEKT trial. Thromb Res 2003;109:101–8. 26. Ignjatovic V, Furmedge J, Newall F, et al. Age-related differences in heparin response. Thromb Res 2006;118:741–5. 27. Andrew M, Mitchell L, Vegh P, Ofosu F. Thrombin regulation in children differs from adults in the absence and presence of heparin. Thromb Haemost 1994;72:836–42. 28. Basu D, Gallus A, Hirsh J, Cade J. A prospective study of the value of monitoring heparin treatment with the activated partial thromboplastin time. N Engl J Med 1972;287:324–7. 29. Hull RD, Raskob GE, Hirsh J, et al. Continuous intravenous heparin compared with intermittent subcutaneous heparin in the initial treatment of proximalvein thrombosis. N Engl J Med 1986;315:1109–14. 30. Ignjatovic V, Summerhayes R, Than J, Gan A, Monagle P. Therapeutic range for unfractionated heparin therapy: age-related differences in response in children. J Thromb Haemost 2006;4:2280–2. 31. Newall F, Ignjatovic V, Summerhayes R, et al. In vivo age dependency of unfractionated heparin in infants and children. Thromb Res 2009;123: 710–4. 32. Ignjatovic V, Summerhayes R, Gan A, et al. Monitoring unfractionated heparin (UFH) therapy: which anti-factor Xa assay is appropriate? Thromb Res 2007;120:347–51. 33. Schedin-Weiss S, Richard B, Hjelm R, Olson ST. Antiangiogenic forms of antithrombin specifically bind to the anticoagulant heparin sequence. Biochemistry 2008;47:13610–9. 34. Akl EA, van Doormaal FF, Barba M, et al. Parenteral anticoagulation for prolonging survival in patients with cancer who have no other indication for anticoagulation. Cochrane Database Syst Rev:CD006652. 35. Adcock DM, Fink LM, Marlar RA, Cavallo F, Zangari M. The hemostatic system and malignancy. Clin Lymph Myeloma 2008;8:230–6. 36. Wojtukiewicz MZ, Sierko E, Rak J. Contribution of the hemostatic system to angiogenesis in cancer. Semin Thromb Hemost 2004;30:5–20. 37. Niessen RW, Lamping RJ, Peters M, Lamers WH, Sturk A. Fetal and neonatal development of antithrombin III plasma activity and liver messenger RNA levels in sheep. Pediatr Res 1996;39:685–91. 38. Schmidt B, Gillie P, Mitchell L, Andrew M, Caco C, Roberts R. A placebocontrolled randomized trial of antithrombin therapy in neonatal respiratory distress syndrome. Am J Respir Crit Care Med 1998;158:470–6.
Contents lists available at ScienceDirect
Blood Reviews journal homepage: www.elsevier.com/locate/blre
REVIEW
Hemostasis in neonates and children: Pitfalls and dilemmas Paul Monagle a,*, Vera Ignjatovic b,1, Helen Savoia c,2 a
Department of Clinical Haematology, Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Flemington Rd. Parkville, Victoria 3052, Australia Haematology Research, Murdoch Children’s Research Institute, Royal Children’s Hospital, Flemington Rd. Parkville, Victoria 3052, Australia c Laboratory Services, Royal Children’s Hospital, Flemington Rd. Parkville, Victoria 3052, Australia b
a r t i c l e Keywords: Hemostasis Neonate Child Development
i n f o
s u m m a r y Developmental Hemostasis refers to the age-related changes in the coagulation system that are most marked during neonatal life and childhood. An understanding of these changes is crucial to the accurate diagnosis of hemostatic abnormalities in neonates and children. This paper explains the current understanding of developmental hemostasis and describes the common pitfalls observed in clinical practice through failure to implement the principles into routine diagnostic work. Finally, there is a brief discussion as to a potential physiological rationale for developmental hemostasis and the implications of this for hemostatic interventions in neonates and children. There remains a need for further study to improve our understanding of the implications of developmental hemostasis in normal growth and development. Ó 2009 Elsevier Ltd. All rights reserved.
Introduction
Developmental hemostasis
Hemostasis is a complex homeostatic system which is critical to maintain healthy life. Disturbances of the hemostatic system account for a large proportion of the non-infectious morbidity and mortality in westernized societies, and are integral to many disease processes including cancer. At a simple level, the critical components of hemostasis are still best described by Virchow’s triad of blood vessel wall, blood composition and blood flow. In considering the non-flow aspects of this equation, hemostasis can be considered as an interaction between the blood vessel wall, coagulation proteins within the plasma, and blood cellular components, predominantly platelets. The hemostatic system interacts with other physiological systems such as wound repair, inflammation and angiogenesis, and the extent and nature of these interactions are still being defined. For the purposes of this paper, the discussions of the hemostatic system will be confined to the coagulation proteins within the plasma. While recognizing the importance of the blood vessel wall and blood cellular interactions, the role of these components are in many ways far less understood at this stage. However, given the nature of clinical laboratory testing, they may also present fewer diagnostic dilemmas for practicing clinicians.
Hemostasis is a dynamic, evolving process that is age-dependent and begins in utero. The evolution of the hemostatic system continues throughout life, however the changes are most marked during childhood, and hence of most clinical relevance during this time. Although evolving, the hemostatic system in healthy fetuses, infants and children must be considered physiologic. The term ‘‘Developmental Hemostasis” was coined in the late 1980s by Dr. Maureen Andrew to describe this phenomenon. Dr. Andrew demonstrated, in a single large cohort study of Canadian children that the concentrations of the majority of coagulation proteins, as measured by functional assays, varied significantly with age. The landmark papers, published in Blood in 1987, 1988 and 1992, not only changed the approach to research related to the coagulation system in children, but also the clinical interpretation of coagulation assays used by many hemostasis laboratories around the world since that time.1–3 However, almost three decades later, there remain many gaps in our understanding of the mechanisms, implications of, and reasons for these observed age-related changes. The evaluation of newborns and children for hemorrhagic or thrombotic complications presents many potential pitfalls and dilemmas that are not encountered in adults. This paper aims to describe the current state of knowledge with respect to developmental hemostasis and then discuss the common pitfalls and strategies to avoid these problems. An understanding of developmental hemostasis in the broadest sense optimizes the prevention, diagnosis, and treatment of hemostatic problems during childhood and undoubtedly provides new insights into the pathophysiology of hemorrhagic and thrombotic complications for all ages.
* Corresponding author. Tel.: +61 3 9345 5914; fax: +61 3 9349 1819. E-mail addresses: [email protected] (P. Monagle), [email protected] (V. Ignjatovic), [email protected] (H. Savoia). 1 Tel.: +61 3 9936 6520; fax: +61 3 8341 6212. 2 Tel.: +61 3 9345 5914; fax: +61 3 9349 1819. 0268-960X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.blre.2009.12.001
64
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
Age-related changes in the coagulation system The fundamental principle underlying developmental hemostasis is that the functional levels of coagulation proteins change in a predictable way with changes in age.4 While the absolute values of these changes are reagent and analyzer dependent, the trends observed are consistent across a number of studies.5 These changes in functional protein levels lead to corresponding changes in global tests of coagulation such as the Activated Partial Thromboplastin Time (APTT). Other global measures of hemostasis may be more or less sensitive to age-related changes. For example, there is little difference in normal thromboelastography (TEG) with age.6 Studies to date have predominantly involved functional assays of the coagulation proteins due to the fact that these assays are commonly used in clinical practice. Whether these changes represent true changes in plasma protein concentrations, or functional changes associated with post-translational protein modifications, or the presence or absence of other as yet unidentified co-factors remains unknown. Previous studies have confirmed that post-translational modifications in coagulation proteins do occur with age. For example, fibrinogen has previously been demonstrated to exist in a ‘‘fetal” form, in cord blood of term infants.7 This ‘‘fetal” fibrinogen was shown to have increased sialic acid content compared to adult fibrinogen, a direct result of post-translational modification. Moreover, in 1970 Witt et al. reported a twofold increase in the phosphorus content of foetal fibrinogen in comparison to that measured in adult fibrinogen.8 These findings were corroborated by a subsequent report of 3–4-fold increase in phosphorus content in fetal fibrinogen compared to adult fibrinogen.9 In addition, specific thrombin clotting times were prolonged in newborns suggesting differences in polymerisation of fibrin from ‘‘fetal” fibrinogen,10 an observation that has lead to claims that fibrinogen in infants is ‘‘dysfunctional”.11 Observations that an increase in sialic acid content of fibrinogen is associated with a decreased rate of fibrin polymerisation while the removal of sialic acid residues leads to increase in polymerisation12 provide a possible explanation for the differences in thrombin clotting times. Definite evidence of the importance of differences in sialic acid content of fibrinogen is provided by observations that sialic residues of fibrinogen directly bind Ca2+.13 This binding leads to a decrease in the intermolecular repulsion between the fibrinogen chains and via that mechanism facilitates fibrin polymerisation.13 Whether such changes exist in older children, and the extent to which similar changes (phosphorylation, glycosylation etc) exist in other hemostatic proteins remains to be determined. The impact of these biochemical changes on routine clinical hemostatic assays, and therefore developmental hemostasis as reflected in routine clinical diagnostic assays is unclear. Table 1 lists major studies that have described age-related changes in coagulation proteins and routine coagulation assays. This paper will not describe all the reported changes in detail, but refers the reader to the papers cited in Table 1. Despite the changes in individual protein levels and in global tests of coagulation, the hemostatic system in neonates and children does not seem disadvantageous compared to the ‘‘normal” coagulation system as measured in adults. There is no data to support either an increased bleeding or thrombotic risk during infancy and childhood for any given stimulus and on the contrary, one could argue that the hemostatic system in neonates and children is protective against bleeding and thrombotic complications compared to adults. This is despite the fact that when considering individual proteins, many proteins exist at levels during stages of infancy that would be associated with disease in adults. For example, normal Antithrombin (AT) levels during the first three months
of life are lower than that observed in many adults with heterozygous AT deficiency and recurrent thrombosis.7 Similarly levels of Factor IX may be below 30% of adult levels in early life and yet these children have no evidence of a bleeding phenotype. Prolongation of global tests of coagulation, such as the APTT, are also not associated with any increase in bleeding, despite the APTT being considerably longer compared to that observed in many adults with clinically relevant VWD.7 The absolute levels of hemostatic proteins as measured by the currently available functional assays are clearly not the key factor in determining clinical phenotype, and there remains much research to be performed to better understand the relationships between levels of hemostatic proteins and functional outcomes of the hemostatic system.
Pitfalls and dilemmas interpreting hemostatic testing in neonates and children Sample acquisition Pre-analytical variables are known to be important in coagulation testing for any age-group. However, nowhere is this more problematic than in neonates and children. Obtaining sufficient blood via clean venepuncture can be extremely challenging in infants. Frequently bloods are obtained via central venous or arterial access devices, where heparin contamination or sample activation are common place. Blood may be collected into a pre-heparinised syringe prior to being transferred to a citrate tube. Further, many neonates will have a relative polycythaemia, meaning that the desired citrate-to-blood ratio (9:1) may not have been achieved, and collection into an adjusted tube is required. Alternatively, collection tubes are frequently under-filled, which is likely not significant for a chemistry panel or a full blood examination, but will ultimately produce inaccurate coagulation results if the citrateto-blood ratio has once again been altered.
Defining reference ranges A number of studies have confirmed that age-related reference ranges for coagulation assays in the healthy population are analyzer and reagent dependent (see Table 1). Yet, very few clinical laboratories have adequately determined age-related reference ranges specific to their analyzer–reagent combination. This deficiency is most likely due to the cost and ethical difficulties in collecting blood from otherwise healthy children to enable reference range determination. Thus, many laboratories compare their pediatric results to either their adult reference range or to published pediatric reference ranges. Neither of these options is satisfactory in terms of the highest standard of clinical care of the patients. Take for example a 4 year-old child who has an APTT performed, using PTT-A reagent (Diagnostica Stago, France) on an STA- compact analyzer (Diagnostica Stago, France). Ninety-five percent confidence interval for a reference range in this age-group is 33–44 s (rounded to whole seconds). If in contrast, the reagent being used is CK-PREST (Diagnostica Stago, France), even with the same analyzer, the reference range encompassing 95% of the population is reported to be 30–35 s.7 Thus if one was referencing to a published reference range based on CK-PREST analysis, but using PTT-A as your reagent, then over 30% of children (assuming normal distribution) in this age-group would be labeled as having prolonged APTT when in fact they were perfectly healthy. Given that the majority of APTT reagents currently in clinical use have no published age-related references ranges, such errors are likely to be frequent. As each new reagent is released onto the market, unless the appropri-
65
P. Monagle et al. / Blood Reviews 24 (2010) 63–68 Table 1 Studies reporting age-related differences in coagulation assays or proteins during childhood. Author
Year
Assays/proteins reported
Age-groups
N
Perlman
1975
PT, TT, APTT, fibrinogen, FDP, platelet count, haematocrit, FV, FVIII, plasminogen, hemoglobin
Healthy infants Small-for-dates infants Post-mature infants
n = 35 n = 26 n = 30
Beverley, D.W. et al.
1984
APTT, FII-VII-X, fibrinogen, a2-antiplasmin, platelet count, MPV, megathrombocyte index, plasminogen
Cord blood Newborns (48 h)
n = 80
Andrew, M. et al.
1987
PT, APTT, TCT, fibrinogen, FII, FV, FVII, FVIII, vWF, FIX, FX, FXI, FXII, PK, HMW-K, FXIIIa, FXIIIb, plasminogen, antithrombin, a2-M, a2-AP, C1E-INH, a1-AT, HCII, protein C, protein S
Day 1 newborn Day 5 newborn Day 30 newborn Day 90 newborn Day 180 newborn Adult
28–75 samples per agegroup
Andrew, M. et al.
1988
PT, APTT, TCT, fibrinogen, FII, FV, FVII, FVIII, vWF, FIX, FX, FXI, FXII, PK, HMW-K, FXIIIa, FXIIIb, plasminogen, antithrombin, a2-M, a2-AP, C1E-INH, a1-AT, HCII, protein C, protein S
Premature newborns (30–36 weeks gestation) Day 1 Day 5 Day 30 Day 90 Day 180
23–67 samples per agegroup
Andrew, M. et al.
1992
PT/INR, APTT, bleeding time, fibrinogen, FII, FV, FVII, FVIII, vWF, FIX, FX, FXI, FXII, PK, HMW-K, FXIIIa, FXIIIs, plasminogen, TPA, PAI, antithrombin, a2-M, a2-AP, C1E-INH, a1-AT, HCII, protein C, protein S (total and free)
1–5 years 6–10 years 11–16 years Adults
20–50 samples per agegroup
Reverdiau-Moalic, P. et al.
1996
PT/INR, APTT TCT, FI, FII, FVII, FVII, FIX, FX, FV, FVIII, FXI, FXII, PK, HMWK, AT, HCII, TFPI, protein C (Ag, Act), protein S (free and total), C4b-BP
Fetuses 19–23 24–29 30–38 weeks gestation Newborns (immediately after delivery) Adults
Cargo, M.D. et al.
2002
PFA100 Hb platelet count
Neonates Children Adults
n = 17 n = 57 n = 31
Salonvaara, M. et al.
2003
FII, FV, FVII, FX, APTT, PT/INR, platelet count
Premature infants 24–27 weeks 28–20 weeks 31–33 weeks 34–36 weeks
n = 21 n = 25 n = 34 n = 45
Flanders, M. et al.
2004
PT, APTT, FVIII, FIX, FXI, antithrombin, RCF, vWF, protein C, protein S
7–9 years 10–11 years 12–13 years 14–15 years 16–17 years Adults
124 per age-group
Monagle, P. et al.
2006
APTT (4 reagents), PT/INR, fibrinogen, TCT, FII, FV, FVII, FVIII, FIX, FX, FXI, FXII, antithrombin, protein C, protein S, D-dimers, TFPI (free and total), endogenous thrombin potential
Day 1 Day 3
Minimum 20 samples per age-group
n = 20 n = 22 n = 22 n = 60 n = 40
<1 years 1–5 years 6–10 years 11–16 years Adults Chan, K.L. et al.
2007
Thromboelastography (TEG) [R. K. a, MA, LY30]
<1 years 1–5 years 6–10 years 11–16 years Adults
n = 24 n = 24 n = 26 n = 26 n = 25
Sosothikul, D. et al.
2007
PT, APTT, fibrinogen, TAT, PC:Ac, TF, FVIIa, sTM, vWF (Ag & RCo), D-dimer, tPA, PAI-1, TAFI
1–5 years 6–10 years 11–18 years Adults
n = 19 n = 26 n = 25 n = 26
Mitsiakos, G. et al.
2008
INR, PT, APTT, fibrinogen, FII, FV, FVII, FVIII, FIX, FX, FXI, FXII, antithrombin, protein C, protein S, APCr, tPA, PAI-1, VWF
Small for growth newborns Appropriate for growth newborns
n = 90 n = 98
Newall, F. et al.
2008
PF4 and Vitronectin
<1 years 1–5 years 6–10 years 11–16 years Adults
15 per age-group
(continued on next page)
66
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
Table 1 (continued) Author
Year
Assays/proteins reported
Age-groups
N
Ries, M. et al.
1997
TAT, F1+2, PAP, D-dimer
1–6 years 7–12 years 13–18 years Adults
20 per age-group
Boos, J. et al.
1989
PIVKA II, FVII, FII, FII:Ag
Day 1, 2, 3 neonates
n = 57 total
ate normative data is generated and published, then children will continue to be classified as either normal or abnormal erroneously. Similar outcome is observed, if an adult based reference range is used. For example, the adult reference range for an APTT using PTTA reagent (Diagnostica Stago, France) and an STA-Compact analyzer is 27–38 s (95% confidence intervals rounded to whole seconds). Once again approximately 30% of children aged 1–5 years would be classified as abnormal if they were compared to this adult reference range, even if their samples were analyzed using the same reagent and analyzer combination. The clinical implications of this type of error are enormous. If one considers the most likely clinical indications for performing an APTT assay in non-hospitalized children, then the true cost of this misclassification starts to become apparent. Children commonly have coagulation studies ordered because they are preoperative and the surgeon is concerned about the past history (‘‘my child bruises easily, or has blood noses”), or because of a positive family history. Erroneous interpretation of an APTT as prolonged, due to the use of an inappropriate reference range, leads to cancellation of surgery, multiple investigations, referral to specialist hematologists and often over-treatment of the child during the procedure. Similarly, children often have coagulation studies because of a positive family history of a bleeding disorder. Misclassification of the child due to inappropriate reference ranges can lead to a child being labeled as having a bleeding or thrombotic disorder when in reality they do not. This can have significant lifestyle and other implications for the child and their family which can be problematic for many years to come. Reversing an already made ‘‘diagnosis” is never easy. Finally, children thought to be at risk from physical or mental abuse will often have coagulation studies as part of their workup if they present with unusual bruising or injuries. Trying to explain in a court of law that you do not believe a child has a bleeding disorder, and that the bruises have truly been a result of child abuse, when the APTT or other coagulation assay results lie outside the reference range quoted by your laboratory is very difficult and may contribute to a miscarriage of justice and further danger to the child. In summary, the lack of use of appropriate reference ranges remains one of the most common pitfalls in the interpretation of pediatric coagulation assays. This practice contributes to inappropriate investigation of children, as well as erroneous diagnosis and management of often healthy children. All laboratories which report coagulation studies in neonates and children must ensure that they are using appropriate age-related reference ranges that are specific for their analyzer–reagent combination. Diagnosing disease Of course, having a value for a hemostatic test, which is outside of the defined 95% confidence intervals for age does not necessarily define disease. Herein exists another dilemma for assessment of hemostasis in neonates and children. For example, if the lower limit of the reference range for plasma levels of factor VIII in a child is 59% based on a population study, and a patient of appropriate age presents who, when using the same analyzer and reagent, has a plasma level for factor VIII of 50%, then it is unlikely that the pa-
tient has even mild hemophilia. This is because by definition, we define hemophilia by arbitrary plasma levels (<30% for mild hemophilia), or at least plasma levels that are thought to be associated with increased risk of bleeding. In the absence of genetic studies definition of heterozygote states remains difficult. The dilemma is that for the laboratory tests to be clinically useful, we need to understand the level at which results define clinically relevant disease probably more so than the levels at which the results are outside the 95% confidence limits of the healthy population. To date, there have been no specific studies in children which address this issue. Correlation of phenotype and genotype, as well as longitudinal studies mapping the hemostatic protein changes in a cohort of patients in conjunction with their clinical outcomes would be most helpful. The issue of results being outside the reference range versus being indicative of clinical disease is most problematic when considering the procoagulant proteins. One must always consider that a positive diagnosis of a hemostatic disorder should include the presence of a family history, a positive clinical phenotype, and reproducible abnormal laboratory results. The problem is that the clinical phenotype is much more difficult to define given the infrequency of thrombotic events; the family history is often equivocal in terms of the role of clinical precipitants of thrombosis, and yet the labeling of children as having a thrombophilic condition can have significant impact. Frequently clinicians are asked to advise on the need for thrombopropylaxis related to clinical interventions, oral contraceptive pill use, or prolonged travel in children in whom the presence or absence of a true thrombophilic condition seems almost impossible to determine. New diagnostic assays The quest for improved clinically predictive value from diagnostic assays is never ending. Over recent years, thrombin generation assays of various types have been promoted as the ideal coagulation assays. In theory, based on our current understanding of the hemostatic system, this seems logical, as thrombin must be viewed as the key enzyme within the hemostatic system and hence understanding the level of thrombin activity should be most helpful in determining whether the hemostatic system is primed for clotting or susceptible to bleeding. The Endogenous Thrombin Potential (ETP) provides an in vitro measure of the overall ability to generate thrombin, which in turn converts fibrinogen to fibrin. As stated, the ETP therefore provides perhaps the best assessment of global hemostasis.14 Currently available chromogenic and fluorogenic substrates available for ETP measurement are cleaved by both free (active) and alpha-2-macroglobulin-bound (inactive) thrombin, leading to an overestimation of thrombin generation. The currently available commercial methods, such as automated fluorogenic assays, account for this using a mathematical algorithm which assumes the likely contribution of alpha-2-macroglobulin to the ETP.15–17 Studies of developmental hemostasis have documented physiological variation in alpha-2-macroglobulin levels with age, where increased levels are observed throughout infancy and childhood compared to adults.1–3,5 The relative contribution of alpha-2macroglobulin bound to and inhibiting thrombin is significant and has been reported to vary from 16% to 64% across the pediatric age-
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
range, while being approximately 7% in adults.18 All methods of thrombin generation measurement must therefore account for this significant, inactive thrombin complex. In particular, the variation with age must be accounted for to prevent over-estimation in the younger populations, where the alpha-2-macroglobulin component plays a significant role. Automated methods that do not directly measure the alpha-2-macroglobulin bound component, but instead rely on algorithms derived from adult populations, are unlikely to account for the significant developmental variation in alpha-2-macroglobulin across the age. Thus, methods which depend on preset algorithms are unlikely to provide accurate results in children of all ages.18 This example demonstrates that the introduction of all new assays into clinical use in children should be based on specific clinical studies in children, confirming the clinical predictive value of the assays, or at the very least should be based on a thorough understanding of the developmental changes in the hemostatic system. Use of clinical algorithms As discussed previously, the use of diagnostic assays, the results of which depend on mathematical calculations based on assumptions formulated in adult testing, are unlikely to be accurate in children. The same may be said of clinical decision making algorithms. Risk stratification for treatment or prophylaxis often depends on clinical algorithms which involve the results of laboratory assays. For example, algorithms which utilize the D-dimer in determining the likelihood of pulmonary embolism,19 Such algorithms have rarely been clinically tested in children, and the only studies addressing this issue have reported that the algorithms are not valid in children.20 The reasons for the lack of validity may include the age-related variation in the normal hemostatic system, uncertainties about the ability to truly diagnose asymptomatic hemostatic disease and probably the pathophysiological differences in bleeding and thrombosis development. However, the evidence to date would suggest that algorithms used in adults are unlikely to be valid in children without modifications, that need to be specifically tested in the pediatric population. This poses a constant dilemma for clinicians, who in the absence of specific pediatric data, often look to adult practice for guidance, and yet the basic principles on which the guidance is based is likely flawed. There remains an ongoing need for pediatric-specific studies to develop appropriate clinical algorithms. Monitoring anticoagulation in children Developmental hemostasis likely impacts on anticoagulant therapy in numerous ways, much of which is beyond the scope of this review. However, an understanding of the impact of developmental hemostasis on laboratory monitoring of anticoagulation is important. There have been no studies to determine target therapeutic ranges for any anticoagulant drug in children based on the gold standard of clinical outcome. The only studies performed to date have reported the doses required in children to achieve the therapeutic ranges determined for given indications in adults.21– 25 The likelihood that these therapeutic ranges are ideal in children seems small. First, given that the reference ranges for coagulation assays in healthy children are different to those in healthy adults, then, for example in the case of APTT monitoring of unfractionated heparin, the incremental increase in APTT to achieve the therapeutic range is clearly different.26 Second, the impact on thrombin generation for identical doses of anticoagulants in children versus adults has been shown to be different,27 suggesting that the target therapeutic ranges will likewise be different. Third, and perhaps most important, adult target therapeutic ranges for anticoagulation are based on clinical studies confirming that the target range
67
provides the optimal therapeutic benefit for the least risk (of bleeding).28,29 The different biology of thrombosis (related in part to developmental hemostasis) and the different risk of bleeding (related to the clinical state of the children) make it unreasonable to assume the target ranges will be identical for such vastly different patient populations. In addition, there is considerable data to show that there is a lack correlation of related monitoring assays, for example APTT versus anti-factor Xa assays for monitoring unfractionated heparin in children compared to adults and that the ability to distinguish ‘‘therapeutic” from ‘‘non-therapeutic” is reduced.26,30,31 Finally, the inter-assay variability seems to be exaggerated in children, and this can have significant clinical implications.32
The role of developmental hemostasis in normal physiology Perhaps the most intriguing aspect of developmental hemostasis is trying to understand the rationale for such marked age-related changes. The neonatal hemostatic system is arguably more protective against stimuli that might cause bleeding or thrombosis in adults. Hence, one could view the neonatal hemostatic system as the ideal system, with the effects of aging leading to deterioration in the system from an early age in healthy individuals. An alternative view is that the hemostatic system is so dramatically different in neonates and children for reasons unrelated to hemostasis. There is increasing evidence that the proteins involved in the hemostatic system have multiple functions in multiple physiological systems within the body, such as angiogenesis, inflammation and wound repair. One theory is that it is these systems that drive the developmental changes in the hemostatic system, which are therefore seen as necessary compensations to allow normal hemostatic function. Evidence for this theory is mounting, and one example of this is found in studies related to Antithrombin (AT). Recently, AT has been shown to have potent anti-angiogenic properties as well as that of being an important anticoagulant.33 AT has been shown to down regulate several pro-angiogenic genes and upregulate a number of anti-angiogenic genes. The anti-angiogenic forms of AT are known to include heparin binding sites, and heparin is shown to potentiate this effect.33 This is one of the postulated mechanisms in the positive effect of heparinoids on cancer survival which is independent of the antithrombotic effect.34–36 AT levels are naturally reduced in newborns to less than 50% of the levels observed in adults and then increase to approach adult levels by approximately six months of age.5 Early evidence suggests that there is a difference in the balance of isoforms of AT in newborns compared to adults. One postulated reason for the decreased levels of AT (and the altered balance of isoforms) observed in neonates is related to the role of this protein in angiogenesis.37 Fetal and early neonatal life is a time of prolific angiogenesis, much more so than any later stage of life, and given the known antiangiogenic properties of AT, then low levels of this protein are likely beneficial for healthy development. Hence, AT replacement therapy during neonatal life may well be deleterious by altering the normal balance of angiogenesis. In the only published randomized trial of AT replacement therapy in newborn infants (as a treatment for lung disease of prematurity), there were seven (11.5%) deaths in the AT-treated group and three (4.9%) deaths in the placebo (no treatment) group.38 A similar trial (reported only in abstract format) also reported a similar trend in the setting of a clinical trial. However, the mechanism for this finding was not able to be formally established.38 This is clearly an area that requires more research. The broader implications of manipulating the hemostatic system in children and infants is vital if we are to avoid non-hematological complications in these patients.
68
P. Monagle et al. / Blood Reviews 24 (2010) 63–68
Conclusion In conclusion, developmental hemostasis is an important concept that has specific implications in the diagnosis of hemostatic disorders in children. Over-diagnosis and missed-diagnosis are both common when age-appropriate, analyzer and reagent specific reference ranges are not used. Further, as new assays are developed, or clinical algorithms utilized without a clear understanding of the impact of developmental hemostasis on their methodology and underlying principles, then the clinical utility of such assays/ algorithms in children will be uncertain. Finally, the rationale for the observed age-related changes in the hemostatic system remains unknown, and yet this would seem crucial to the determination of risk-benefit ratios for therapeutic manipulations of the hemostatic system in neonates and children. There remains an urgent need for more research into this crucial aspect of normal human development. Practice points Sample integrity is a major problem in pediatric coagulation studies. Attention to detail and use of repeat samples are important to avoid erroneous results. Age appropriate (analyzer and reagent specific) reference ranges are critical to the accurate diagnosis and management of coagulation disorders in children. Defining hemostatic disease requires more than a laboratory parameter outside of a population based reference range. Clinical algorithms used in adult practice need to be specifically tested in pediatric settings before they can be applied to neonates and children.
Research agenda The physiological causes responsible for the observed agerelated changes in coagulation proteins remain to be elucidated and are of tremendous biological significance. The development of pediatric-specific algorithms to guide clinical management requires specific clinical outcome studies in children. New assays need to be validated in children, in light of our understanding of developmental hemostasis.
Conflict of interest Paul Monagle and Vera Ignjatovic have received research funding from Diagnostico Stago, Roche, and Bayer. References 1. Andrew M, Paes B, Milner R, et al. Development of the human coagulation system in the healthy premature infant. Blood 1988;72:1651–7. 2. Andrew M, Paes B, Milner R, et al. Development of the human coagulation system in the full-term infant. Blood 1987;70:165–72. 3. Andrew M, Vegh P, Johnston M, Bowker J, Ofosu F, Mitchell L. Maturation of the hemostatic system during childhood. Blood 1992;80:1998–2005. 4. Andrew M, Paes B, Johnston M. Development of the hemostatic system in the neonate and young infant. Am J Pediatr Hematol Oncol 1990;12:95–104. 5. Monagle P, Barnes C, Ignjatovic V, et al. Developmental haemostasis. Impact for clinical haemostasis laboratories. Thromb Haemost 2006;95:362–72. 6. Chan KL, Summerhayes RG, Ignjatovic V, Horton SB, Monagle PT. Reference values for kaolin-activated thromboelastography in healthy children. Anesth Analg 2007;105:1610–3. [table of contents]. 7. Monagle P, Hagstrom J. Developmental haemostasis. In: Polin, editor. Fetal and neonatal physiology. 3rd ed. Elsevier: St. Louis; 2003. p. 1435–47. 8. Witt I, Muller H. Phosphorus and hexose content of human foetal fibrinogen. Biochim Biophys Acta 1970;221:402–4.
9. Hamulyak K, Nieuwenhuizen W, Devilee PP, Hemker HC. Reevaluation of some properties of fibrinogen, purified from cord blood of normal newborns. Thromb Res 1983;32:301–10. 10. Witt I, Muller H, Kunzer W. Evidence for the existence of foetal fibrinogen. Thromb Diath Haemorrh 1969;22:101–9. 11. Miller BE, Tosone SR, Guzzetta NA, Miller JLKK, Brosius KK. Fibrinogen in children undergoing cardiac surgery: is it effective? Anesth Analg 2004;99:1341–6. 12. Martinez J, MacDonald KA, Palascak JE. The role of sialic acid in the dysfibrinogenemia associated with liver disease: distribution of sialic acid on the constituent chains. Blood 1983;61:1196–202. 13. Dang CV, Shin CK, Bell WR, Nagaswami C, Weisel JW. Fibrinogen sialic acid residues are low affinity calcium binding sites that influence fibrin assembly. J Biol Chem 1989;264:15104–8. 14. Al Dieri R, Peyvandi F, Santagostino E, et al. The thrombogram in rare inherited coagulation disorders: its relation to clinical bleeding. Throm Haemost 2002;88:576–82. 15. Hemker HC, Beguin S. Thrombin generation in plasma: its assessment via the endogenous thrombin potential. Thromb Haemost 1995;74:134–8. 16. Hemker HC, Giesen P, Al Dieri R, Regnault V, de Smedt E, Wagenvoord R. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol Haemost Thromb 2003;33:4–15. 17. Hemker HC, Wielders S, Kessels H, Beguin S. Continuous registration of thrombin generation in plasma, its use for the determination of the thrombin potential. Thromb Haemost 1993;70:617–24. 18. Ignjatovic V, Greenway A, Summerhayes R, Monagle P. Thrombin generation: the functional role of alpha-2-macroglobulin and influence of developmental haemostasis. Br J Haematol 2007;138:366–8. 19. Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. Ann Intern Med 2001;135:98–107. 20. Biss TT, Brandao LR, Kahr WH, Chan AK, Williams S. Clinical probability score and D-dimer estimation lack utility in the diagnosis of childhood pulmonary embolism. J Thromb Haemost 2009;7:1633–8. 21. Bauman ME, Belletrutti MJ, Bajzar L, et al. Evaluation of enoxaparin dosing requirements in infants and children. Better dosing to achieve therapeutic levels. Thromb Haemost 2009;101:86–92. 22. Dix D, Andrew M, Marzinotto V, et al. The use of low molecular weight heparin in pediatric patients: a prospective cohort study. J Pediatr 2000;136:439–45. 23. Kuhle S, Massicotte P, Dinyari M, et al. Dose-finding and pharmacokinetics of therapeutic doses of tinzaparin in pediatric patients with thromboembolic events. Thromb Haemost 2005;94:1164–71. 24. Massicotte P, Adams M, Marzinotto V, Brooker LA, Andrew M. Low-molecularweight heparin in pediatric patients with thrombotic disease: a dose finding study. J Pediatr 1996;128:313–8. 25. Massicotte P, Julian JA, Gent M, et al. An open-label randomized controlled trial of low molecular weight heparin for the prevention of central venous linerelated thrombotic complications in children: the PROTEKT trial. Thromb Res 2003;109:101–8. 26. Ignjatovic V, Furmedge J, Newall F, et al. Age-related differences in heparin response. Thromb Res 2006;118:741–5. 27. Andrew M, Mitchell L, Vegh P, Ofosu F. Thrombin regulation in children differs from adults in the absence and presence of heparin. Thromb Haemost 1994;72:836–42. 28. Basu D, Gallus A, Hirsh J, Cade J. A prospective study of the value of monitoring heparin treatment with the activated partial thromboplastin time. N Engl J Med 1972;287:324–7. 29. Hull RD, Raskob GE, Hirsh J, et al. Continuous intravenous heparin compared with intermittent subcutaneous heparin in the initial treatment of proximalvein thrombosis. N Engl J Med 1986;315:1109–14. 30. Ignjatovic V, Summerhayes R, Than J, Gan A, Monagle P. Therapeutic range for unfractionated heparin therapy: age-related differences in response in children. J Thromb Haemost 2006;4:2280–2. 31. Newall F, Ignjatovic V, Summerhayes R, et al. In vivo age dependency of unfractionated heparin in infants and children. Thromb Res 2009;123: 710–4. 32. Ignjatovic V, Summerhayes R, Gan A, et al. Monitoring unfractionated heparin (UFH) therapy: which anti-factor Xa assay is appropriate? Thromb Res 2007;120:347–51. 33. Schedin-Weiss S, Richard B, Hjelm R, Olson ST. Antiangiogenic forms of antithrombin specifically bind to the anticoagulant heparin sequence. Biochemistry 2008;47:13610–9. 34. Akl EA, van Doormaal FF, Barba M, et al. Parenteral anticoagulation for prolonging survival in patients with cancer who have no other indication for anticoagulation. Cochrane Database Syst Rev:CD006652. 35. Adcock DM, Fink LM, Marlar RA, Cavallo F, Zangari M. The hemostatic system and malignancy. Clin Lymph Myeloma 2008;8:230–6. 36. Wojtukiewicz MZ, Sierko E, Rak J. Contribution of the hemostatic system to angiogenesis in cancer. Semin Thromb Hemost 2004;30:5–20. 37. Niessen RW, Lamping RJ, Peters M, Lamers WH, Sturk A. Fetal and neonatal development of antithrombin III plasma activity and liver messenger RNA levels in sheep. Pediatr Res 1996;39:685–91. 38. Schmidt B, Gillie P, Mitchell L, Andrew M, Caco C, Roberts R. A placebocontrolled randomized trial of antithrombin therapy in neonatal respiratory distress syndrome. Am J Respir Crit Care Med 1998;158:470–6.