OFFICIAL COMMUNICATION OF THE SSC
Classification of rare bleeding disorders (RBDs) based on the association between coagulant factor activity and clinical bleeding severity
Present address: Flora Di Michele is currently in the Division of Blood Diseases and Resources at the National Heart, Lung, and Blood Institute, Bethesda, MD, USA.
Flora Peyvandi, U.O.S. Dipartimentale per la Diagnosi e la Terapia delle Coagulopatie, A. Bianchi Bonomi Hemophilia and Thrombosis Center, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Università degli Studi di Milano and Luigi Villa Foundation, Milan, Italy.
Tel.: +39 2 55035414; fax: +39 2 54100125.
Rare bleeding disorders (RBDs) include the inherited deficiencies of coagulation factors, fibrinogen, factor (F)II, FV, FV + FVIII, FVII, FX, FXI and FXIII, and are usually transmitted as autosomal recessive disorders . Due to their rarity, they sometimes present significant challenges in diagnosis and treatment . The development of guidelines for classification and treatment of these disorders has been hampered by a lack of sufficient knowledge about epidemiology and clinical outcomes, the difficulty in recognizing affected patients and collecting longitudinal clinical data, the limits of laboratory assays, and a lack of consensus concerning the criteria by which these disorders are classified. A collaboration among experts on RBDs was therefore proposed.
Over the past 5 years, the Rare Bleeding Disorders Working Group, under the umbrella of the Factor VIII and Factor IX Scientific and Standardisation Committee of the International Society on Thrombosis and Haemostasis, has started to focus on the need for comprehensive data collection regarding laboratory phenotype and clinical manifestations in RBDs. Different networks and registries, established in Europe (European network of Rare Bleeding Disorders – (EN-RBD)) , the United Kingdom (United Kingdom Haemophilia Centre Doctors’ Organisation registry (UKHCDO)) , the USA (The North American Rare Bleeding Disorders Registry (NARBDR)  and the American Thrombosis Hemostasis Network (ATHN) Rare Coagulation Disorder database (http://www.athn.org)) and India  started to collaborate with an initial task of understanding what data are currently available through each system. Here, we report the results of a preliminary review of available data that explored the association between residual plasma coagulant factor activity and the clinical bleeding profile for each of the RBDs. Combined vitamin K-dependent clotting factor deficiencies are not included in this analysis. The results are based on two different sets of data.
- 1 A detailed review of the available literature. We searched Medline and PubMed for articles published in English between January 1990 and March 2012, with the following search terms: ‘clinical manifestations’ OR ‘bleeding symptoms’ OR ‘genotype phenotype’ OR ‘outcomes’ in combination with the specific type of deficiency OR ‘rare bleeding disorder’. We excluded articles on the basis of their titles or abstracts and included remaining articles and those identified from reference lists of relevant reports that reported data on the association between the laboratory phenotype and clinical bleeding characteristics for at least five patients. We identified 51 relevant original articles (Table S1) and 39 review articles (Table S2).
- 2 Overview of data on the association between the laboratory phenotype and clinical bleeding characteristics from the EN-RBD, UKHCDO, NARBDR and Indian registries for a total of 4359 patients affected with RBDs (592, 3192, 294 and 281, respectively).
The following classification systems were identified in the original resources and used to interpret the association between the residual plasma coagulant factor activity and the clinical bleeding severity.
- 1 Residual plasma coagulant factor activity was classified as mild, moderate or severe in the UKHCDO, NARBDR and Indian registries as described in Table 1. Classification of clinical bleeding severity was based on site and frequency of bleeding.
- 2 Residual plasma coagulant factor activity was considered as a continuous variable in the EN-RBD, while clinical bleeding severity was classified as described in Table 2.
- 3 For data retrieved from the literature (Tables S1 and S2), residual plasma coagulant factor activity was classified differently by different references, and these are described in Tables S1 and S2. For clinical bleeding severity, we relied on the authors’ original classification of severity, and in cases where that was not provided we used the classification system described in Table 2.
Table 1. Laboratory phenotype classifications according to the four registries, compared with the laboratory classification of hemophilia
|UKHCDO*||< 5% (subdivided into 0 and 1–5%) |
Fibrinogen: < 0.1 g L−1
Fibrinogen: 0.1–0.5 g L−1
|> 10% (subdivided into 10–20% and > 20%) |
Fibrinogen: > 0.5 g L−1
Afibrinogenemia: < 0.5 g L−1
FXIII: abnormal clot lysis assay
Hypofibrinogenemia: 0.5 g L−1 - LLN (lower limit of normal range in local)
|INDIA‡||Afibrinogenemia: no detectable clot |
FV, FVII, and FX: < 1%
FXI: < 2%
FXIII: abnormal clot lysis assay
|Hypofibrinogenemia: < 1 g L−1||Dysfibrinogenemia: > 1 g L−1|
FV + FVIII: < 40–45%§
| Hemophilia ||< 1%||1–5%||> 5%|
Table 2. Assigned categories of clinical bleeding severity in the EN-RBD 
|Asymptomatic||No documented bleeding episodes|
|Grade I bleeding||Bleeding that occurred after trauma or drug ingestion (antiplatelet or anticoagulant therapy)|
|Grade II bleeding|| Spontaneous minor bleeding: bruising, ecchymosis, minor wounds, oral cavity bleeding, epistaxis and menorrhagia|
|Grade III bleeding|| Spontaneous major bleeding: hematomas*, hemarthrosis, CNS, GI and umbilical cord bleeding|
Based on the interpretation of published data and the experiences from the networks and registries, it is evident that:
- 1 there is a heterogeneous association between residual plasma coagulant factor activity and clinical bleeding severity in different RBDs (these data indicate that a more detailed evaluation on each single factor deficiency is required for future planning of optimal diagnosis and management); and
- 2 it is not appropriate to use a single criterion of classification for all types of RBDs.
Despite the heterogeneity of the data available in the literature or collected through the different networks and registries, a general picture of each single deficiency is reported in the following sections.
Information from the literature (Tables S1 and S2) shows that patients with afibrinogenemia have a bleeding tendency of variable severity. The typical symptoms include umbilical cord, mucosal, gastrointestinal tract (GI), genitourinary or central nervous system (CNS) bleeding. Other relatively frequent symptoms include hemarthroses and hematomas. First trimester abortion is common in afibrinogenemic women. Both arterial and venous thromboembolic complications were also recorded in some cases. The bleeding pattern of patients with hypofbrinogenemia is similar but appears milder; however, patients may bleed when exposed to trauma (Tables S1 and S2). The bleeding tendency of patients with dysfibrinogenemia is unpredictable, and hemorrhages occur often after trauma or surgery or in the post-partum period. Thrombosis may also occur. Patients with a fibrinogen level around 1.0 g L−1 are usually asymptomatic (Tables S1 and S2).
Despite different classifications used in the four registries, there is general agreement on clinical manifestations and a good association between laboratory phenotype and clinical severity (and also age at diagnosis) was observed in patients affected with fibrinogen deficiency. The lack of standardization of the available assays makes it difficult to reach conclusions regarding the minimal amount of fibrinogen that could significantly reduce bleeding symptoms. However, EN-RBD data show that there is a strong association between fibrinogen level and clinical bleeding severity, with patients having undetectable levels being more likely to experience major spontaneous bleeding while those with levels > 0.1 g L−1 are mostly asymptomatic.
It is to be mentioned that sometimes the quantification of a low fibrinogen concentration using the Clauss or PT-derived methods, routinely used in the clinical laboratories to measure plasma levels of the protein, may vary significantly and this phenomenon could generate uncertainty in assigning the severity grade of the hypofibrinogenemia. The relative mismatch between the levels measured by different assays and clinical severity of the hypofibrinogenemia may stem also from this phenomenon. The recommendation should be that the Clauss method remains the reference method because it can be performed on all instruments (not the case for the PT-derived method).
Due to the small number of reported cases, any association between bleeding manifestations and residual FII coagulant activity in plasma is very difficult to ascertain. However, FII coagulant activity < 10% is usually reported to be associated with severe bleeding manifestations (Tables S1 and S2), and only a few cases with < 5% of coagulant activity associated with a low level of antigen (type II deficiency) have been reported (Tables S1 and S2).
In the NARBDR, a weak direct correlation between laboratory severity and the age at first bleed was noted. Furthermore, an inverse correlation between laboratory severity and mean lifetime number of trauma-induced and musculoskeletal bleedings was noted, with both NARBDR and hemophilia classification (Table 1). Significant symptoms reported in Indian patients with FII coagulant level < 20% were prolonged post-injury bleeding, mucosal bleeding and subcutaneous and muscle hematoma, while GI bleeding was seen only in a few cases.
Results of the literature review suggest that the severity of symptoms is variable and correlates poorly with laboratory phenotype (Tables S1 and S2).
Data from the registries confirmed this heterogeneity, showing a weak association between laboratory phenotype and clinical severity (including the age of patients at diagnosis or at first bleed). Analysis of the EN-RBD registry demonstrated that undetectable residual coagulant activity was associated with spontaneous major bleeding, while patients with activity levels > 10% seem to remain asymptomatic.
Data from India showed that the most common symptoms in patients with severe FV deficiency were prolonged bleeding post-injury and mucosal bleeding, followed by hematoma, hemarthrosis and GI bleeding. Intracranial and umbilical cord bleeding were recorded in only 10% of patients with severe deficiency.
Platelets provide a concentrated supply of FV: although most FV is present in plasma, approximately 20–25% of the circulating FV is found within platelet α-granules . After α-granule release upon platelet activation, FV can presumably bind immediately to surface receptors optimizing prothrombinase complex activity . This probably protects some patients with FV deficiency from severe life-threatening bleeding, and may explain the weak association between laboratory phenotype and clinical severity.
Compound FV + FVIII deficiency
According to the literature, patients usually present with a wide range of clinical symptoms, ranging from spontaneous hematomas and mucosal/oral bleeding to post-traumatic bleeding (Tables S1 and S2). Patients with mild or moderate compound FV + VIII deficiency usually experience minor bleeding; however, few patients with major bleeding were also recorded (Tables S1 and S2). The same data have been confirmed by the registries that tracked this disorder, and no strong association between laboratory phenotype and clinical manifestations has been observed. In the EN-RBD, patients with activity levels < 20% were most likely to experience spontaneous major bleeding, while those with levels > 40% remained largely asymptomatic.
Data from the literature suggest that hemorrhagic manifestations cannot be easily predicted by the coagulant activity level, and patients with similar coagulant activities can vary greatly in their bleeding symptoms (Tables S1 and S2).
Similarly to FV deficiency, data from the registries confirmed a poor association between laboratory phenotype and clinical severity. Even for this deficiency, data from the EN-RBD registry showed that < 10% of residual coagulant activity seems to be associated with major spontaneous bleedings. It is to be considered that the choice of the reagents used in FVII assays is really important: the diagnostic efficiency of the FVII coagulant activity assays is highly dependent on the thromboplastin sensitivity. In general, thromboplastin of human origin or recombinant thromboplastins structurally identical to the human type should be used .
The most common symptoms reported in the Indian registry were prolonged bleeding post-injury and mucosal bleeding, followed by hematoma, hemarthrosis and GI bleeding. Intracranial and umbilical cord bleeding were recorded in few patients with severe deficiency.
Data from the literature (Tables S1 and S2) show that patients with undetectable FX coagulant activity have severe bleeding manifestations, such as CNS, GI or umbilical cord bleeding, or recurrent hemarthroses and hematomas. These symptoms are uncommon in patients with moderate and mild deficiency, who usually only manifest mucocutaneous bleeding or remain asymptomatic throughout their lifetime.
Similar results were found after analyzing data from the registries: a strong inverse association between FX coagulant level and bleeding severity was noted. In particular, analysis of the EN-RBD registry demonstrated that patients with activity levels < 10% of FX had a high risk of major spontaneous bleeding, such as CNS and GI bleeding, suggesting this as a target level for prophylaxis. However, patients with levels > 40% remained largely asymptomatic and those having levels between 10 and 40% only suffered minor spontaneous or triggered bleeding.
The literature reports that clinical bleeding tendency appears not to be associated with FXI coagulant activity. Spontaneous hemorrhage is uncommon in patients with FXI activity levels < 15–20%, who, on the contrary, have a high probability of postoperative hemorrhage (Tables S1 and S2). Patients with levels between 20 and 65% of FXI coagulant activity usually remain asymptomatic or have a low risk of postoperative bleeding (Tables S1 and S2). Data from registries that tracked this disorder confirmed this finding and did not show any association between clinical bleeding tendency and residual FXI coagulant activity, both when undetectable or moderately reduced (< 20%).
Patients with undetectable FXIII coagulant activity have severe bleeding manifestations, such as CNS or umbilical cord bleeding, or recurrent hemarthroses and hematomas. Miscarriages are very common in women with severe deficiency. Patients with moderate and mild deficiency can have only mucocutaneous bleeding or may be completely asymptomatic (Tables S1 and S2). Data from the registries indicate that the severity of the FXIII laboratory phenotype correlates well with the age at diagnosis, the clinical symptoms and the amount of replacement therapy required for treatment. However, this association in patients with severe FXIII deficiency (< 5%) is particularly difficult to interpret and may not be accurate, because most of the available data are based on the clot solubility test. This is a qualitative test that is poorly standardized and insensitive in the range of values 1–5% of FXIII coagulant activity. In any case, EN-RBD data showed that patients with undetectable FXIII coagulant activity were most likely to experience major spontaneous bleeding, while those with levels ≥ 30% remained primarily asymptomatic, although with wide confidence intervals (10–50%).
In conclusion, although the sources used in this report may be inadequate to provide a basis for evidence-based recommendations for treatment of patients with RBDs, the collection and interpretation of these data allowed us to reach the following classification on the basis of the observed associations between clinical and laboratory severity.
- 1 Fibrinogen, FII, FX and FXIII deficiencies are RBDs with a strong association between clinical severity and coagulant activity level, with a few exceptions.
- 2 FV and FVII deficiencies are RBDs with a poor association between clinical severity and coagulant activity level.
- 3 FXI deficiency shows no association between clinical severity and coagulant activity level, both when undetectable or moderately reduced (< 20%).
- 4 Compound FV + FVIII deficiency is mainly associated with mild or moderate clinical symptoms and patients rarely experience severe bleeding.
Therefore for most of the RBDs, except FXI deficiency, the following severity classification could be proposed (Table 3).
Table 3. Proposal of the project on RBDs
|Fibrinogen||Undetectable clot||0.1−1g L−1||> 1 g L−1|
|FII||Undetectable activity||≤ 10%||> 10%|
|FV||Undetectable activity||< 10%||≥ 10%|
|FV + FVIII||< 20%||20–40%||> 40%|
|FVII||< 10%||10–20%||> 20%|
|FX||< 10%||10–40%||> 40%|
|FXIII||Undetectable activity||< 30%||≥ 30%|
- 1 Severe deficiency: coagulant activity associated with spontaneous major bleeding.
- 2 Moderate deficiency: coagulant activity associated with mild spontaneous or triggered bleeding.
- 3 Mild deficiency: coagulant activity associated with a mostly asymptomatic disease course.
This classification is based primarily on the EN-RBD data but without significant contradictions from the analyses of the other registries.
It has to be mentioned that a proposed classification by Lapecorella et al.  for FVII deficiency based on symptoms and not on activity levels may be taken into consideration for the disorders with poor association between clinical severity and coagulant activity (Table 4).
Table 4. Severity classification of FVII deficiency, as reported in table 1 of reference 
|Severe||They had at least one of the following symptoms: GI or CNS bleeding or hemarthrosis with or without other bleeds|
|Moderate||Those who had three or more symptoms with the exception of GI CNS bleeding or hemarthrosis|
|Mild||Those who had one or two symptoms with the exception of GI CNS bleeding or hemarthrosis|
The antigen level is usually not mandatory for routine diagnostic laboratory investigations, except for fibrinogen and FII deficiencies, in which the antigen measurement could help to differentiate hypo- or dys-prothrombinemia and hypo- or dys-fibrinogenemia, especially because a normal antigen level in plasma could increase the risk of thrombosis.
Despite the large number of patients evaluated in this overview (both from the literature and the aforementioned registries), there is a large heterogeneity in the pre-assigned severity definitions for both coagulant activity and bleeding symptoms. We interpreted the reported associations and derived conclusions from common observations in different resources. Moreover, inherited and acquired hemostatic defects apart from the given disorder have not been ruled out, and thus laboratory versus clinical phenotypes may be misleading.
These limitations underline the future need to develop an accurate data collection tool, available to hemophilia centres around the world. Achieving common definitions between different parties alongside longitudinal data collection would help determine the hemostatic level of each single factor to prevent spontaneous and post-trauma/surgery/delivery hemorrhage for each type of RBD. In fact, such efforts are ongoing and two large longitudinal data collection projects (PRO-RBDD) for patients affected with congenital fibrinogen and FXIII deficiency are underway. Moreover, the heterogeneous nature of the association between the laboratory phenotype and clinical manifestations in some RBDs indicates that other disease modifiers may be in play, which merit further study. It also points out the inadequacy of the available assays to evaluate the minimum residual level of coagulation factor to prevent the risk of bleeding. The assays used to measure plasma levels of the coagulation factor and the choice of the reagents are also really important to consider, because there are usually significant inter-laboratory differences in the results of factor assays. Some consensus on factor assay methodology is important so that values from different laboratories/centres can be compared. This conclusion should pave the way to define the role of other global measurements such as thrombin generation and/or thromboelastography in the definition of hemorrhagic risk. Global assay standardization and application in RBDs should represent the next step to be undertaken by the working group in organising centralized laboratory assessment.
F. Peyvandi designed the study, supervised the data collection and wrote the manuscript; D. Di Michele provided data obtained by the NARBDR and revised the manuscript; P. H. B. Bolton-Maggs provided and analysed data from the UKHCDO registry and revised the manuscript; C. A. Lee revised the progress of this work and critically revised the manuscript; A. Tripodi critically revised the manuscript; A. Srivastava provided data obtained in India, and critically revised the results and the manuscript. All authors approved the final version of the manuscript.
The authors thank: A. Shapiro (Indiana Hemophilia and Thrombosis Centre, Indianapolis, IN, USA) and C. Negrier (Unité d’Hémostase Clinique, Hôpital Edouard Herriot, Lyon, France) for being members of the Project on Consensus Definitions in Rare Bleeding Disorders of the Factor VIII/Factor IX Scientific and Standardisation Committee of the International Society on Thrombosis and Haemostasis; K. M. Musallam, I. Garagiola and A. Cairo (U.O.S. Dipartimentale per la Diagnosi e la Terapia delle Coagulopatie, A. Bianchi Bonomi Hemophilia and Thrombosis Center, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Università degli Studi di Milano and Luigi Villa Foundation, Milan, Italy) for the accurate revision of data available in the scientific literature; R. Palla and M. Menegatti (U.O.S. Dipartimentale per la Diagnosi e la Terapia delle Coagulopatie, A. Bianchi Bonomi Hemophilia and Thrombosis Center, Fondazione IRCCS Cà Granda Ospedale Maggiore Policlinico, Università degli Studi di Milano and Luigi Villa Foundation, Milan, Italy) for performing statistical analysis, and interpretation of the data collected by the EN-RBD network; L. Dewhurst (Administrator Assistant to the UKHCDO database, Ardwick, Manchester, UK); S. Acharya, co-author of the NARBDR and participant in the secondary data analyses of that registry; A. Viswabandya and S. Chandran Nair (Christian Medical College, Vellore, India) for the data from India; F. Bernardi (Department of Biochemistry and Molecular Biology, University of Ferrara, Italy), R. De Cristofaro (Institute of Internal Medicine and Geriatrics and Hemostasis Research Center, Catholic University School of Medicine, Rome, Italy), P. De Moerloose (Division of Angiology and Haemostasis, University Hospital, Geneva, Switzerland), L. Muszbek (Clinical Research Center, University of Debrecen, Medical and Health Science Center, Debrecen, Hungary), D. Perry (Department of Haematology, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK) and U. Selighson (The Amalia Biron Research Instistute of Thrombosis and Hemostasis, Sheba Medical Center, Tel Hashomer, Isreal), for their comments and critical revision of the manuscript; and all the collaborating partners involved in the EN-RBD project (http://www.rbdd.eu/partners.htm):
- 1 Hemofilie Centrum Leuven, Katholieke Universiteit, Leuven, Belgium (K. Peerlink);
- 2 Antwerp University Hospital UZA, Edegem, Belgium (A. Gadisseur);
- 3 Centre for Hemophilia and Thrombosis, Department of Clinical Biochemistry, University Hospital Skejbi, Aarhus, Denmark (J. Ingerslev);
- 4 Hopital Saint Eloi, Montpellier, France (J.F. Schved and M. Giansily–Blaizot) and associated centres;
- 5 Pediatric Hemophilia and Thrombosis Centre, Dr. von Hauner’s Children’s University Hospital, Munich, Germany (C. Bidlingmaier);
- 6 MVZ Labor Duisburg GmbH – Duisburg, Germany (S. Halimeh);
- 7 Haemophillia Center, Haemostasis Unit, Agia Sofia Children’s Hospital, Athens, Greece (H. Platokouki, H. Pergantou);
- 8 First Regional Transfusion and Haemophilia Center, Hippocration Hospital, Athens, Greece (G. Theodossiades);
- 9 ‘Laiko’ General Hospital – 2nd Blood Transfusion Center and Haemophilia Center, Athens, Greece (O. Katsarou);
- 10 Hippoktrateion Hospital, Tessalonike, Greece (V. Economou);
- 11 National Centre for Hereditary Coagulation Disorders, St James’s Hospital, Dublin, Ireland (R. Gilmore);
- 12 Angelo Bianchi Bonomi Hemophilia and Thrombosis Center, Università degli Studi di Milano, Milan, Italy (S.M. Siboni);
- 13 Haemostasis Department and Haemophilia Center, Blood Transfusion Institute of Serbia, Belgrade, Serbia (D. Mikovic);
- 14 National Haemophilia Center, University Children’s Hospital, Ljubljiana, Slovenia (M. Benedik-Dolnicar and L. Kitanovski);
- 15 Department of Pediatric Hematology-Oncology, Cerrahpasa Medical Faculty of Istanbul University, Istanbul, Turkey (T. Celkan and N. Özdemir);
- 16 Oxford Haemophilia & Thrombosis Centre, Nuffield Department of Clinical Medicine, University of Oxford, Oxford, United Kingdom (P. Giangrande);
- 17 Royal Free Hampstead NHS Trust, London, United Kingdom (E.G.D. Tuddenham and P. Morjaria).
Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.