John D. Grainger

Over the last decade, there have been numerous developments and changes in treatment practices for the management of patients with immune thrombocytopenia (ITP). This article is an update of the International Consensus Report published in 2010. A critical review was performed to identify all relevant articles published between 2009 and 2018. An expert panel screened, reviewed, and graded the studies and formulated the updated consensus recommendations based on the new data. The final document provides consensus recommendations on the diagnosis and management of ITP in adults, during pregnancy, and in children, as well as quality-of-life considerations.

Abstract Over the last decade, there have been numerous developments and changes in treatment practices for the management of patients with immune thrombocytopenia (ITP). This article is an update of the International Consensus Report published in 2010. A critical review was performed to identify all relevant articles published between 2009 and 2018. An expert panel screened, reviewed, and graded the studies and formulated the updated consensus recommendations based on the new data. The final document provides consensus recommendations on the diagnosis and management of ITP in adults, during pregnancy, and in children, as well as quality-of-life considerations.

The objective of this document is to guide diagnosis and management of patients with rare coagulation disorders (RCD). This document replaces the 2004 UK Haemophilia Centre Doctors' Organization (UKHCDO) rare coagulation disorders guideline (Bolton-Maggs et al, 2004a). The RCD are here defined as monogenic bleeding disorders caused by deficiency of a soluble coagulation factor or factors, other than von Willebrand disease (VWD), Haemophilia A or Haemophilia B. The RCD described in this document include heritable deficiencies of fibrinogen, prothrombin, factor (F) V, FVII, FX, FXI and FXIII, combined FV and FVIII deficiency and vitamin K-dependent coagulation factor deficiency. RCD are usually caused by recessive inheritance of unique or rare nucleotide variations in the genes encoding the coagulation factors or in proteins necessary for their post-translational processing. RCD are more common in ethnic groups in which consanguineous partnerships are common, because of the higher likelihood of homozygosity. Dysfibrinogenaemia and FXI deficiency may show autosomal dominant or recessive inheritance (Table 1). Heterozygous carriers of variations in other classically 'recessive' RCDs sometimes display bleeding symptoms. 1:1 million (AR) Unknown (AD) 1:1 million (AR) 1:30 000 (AD) The writing group was representative of UK experts in RCD. Evidence was gathered from primary English language publications identified in PubMed from 1990 using the disorder names and synonyms as index terms. Relevant reviews and other guidelines were also searched for informative primary publications. The writing group produced a draft guideline that was revised by consensus by the UKHCDO Advisory Group and the Haemostasis and Thrombosis Task Force of the British Committee for Standards in Haematology (BCSH) and the BCSH Executive Committee. The guideline was then reviewed by a sounding board of 50 members of the British Society for Haematology (BSH) who have commented on its content and applicability in the UK setting. The strength of recommendations and quality of evidence are presented in GRADE format http://www.bcshguidelines.com/BCSH_PROCESS/EVIDENCE_LEVELS_AND_GRADES_OF_RECOMMENDATION/43_GRADE.html. Published descriptions of the RCD have historically comprised case reports or short series. However, initiatives, such as the European Network of Rare Bleeding Disorders (EN-RBD; Peyvandi et al, 2012a), the North American Rare Bleeding Disorders Registry (Acharya et al, 2004) and several disease-specific registries (Herrmann et al, 2006, 2009; Ivaskevicius et al, 2007; Bernardi et al, 2009) have improved understanding of the RCD. This has enabled the EN-RBD, under the auspices of the International Society of Thrombosis and Haemostasis to propose laboratory criteria of disease severity for most RCD (Table 1; Peyvandi et al, 2012a). Despite this progress, the clinical characteristics of many RCD remain incompletely documented and management is informed by open label observational studies and not randomized controlled trials. Therefore, the quality of most evidence considered in this guideline is moderate (B) or low (C) and most recommendations are weak (2). This document is intended to guide factor replacement or other therapies for most clinical scenarios. However, clinicians are expected to modify treatment plans according to the severity of individual bleeds or procedures and to the background bleeding phenotype of each case. Further guidance about laboratory evaluation, selection of therapeutic products and the management of women with RCD, including regional anaesthesia, is provided in previous UKHCDO or BCSH guidelines (Lee et al, 2006; Keeling et al, 2008). In common with other heritable bleeding disorders, the treatment and prevention of bleeding in the RCD requires general measures, such as avoiding high bleeding risk activities, selecting invasive procedures with the minimum bleeding risk and ensuring adequate communication of treatment plans developed by haemophilia centres with appropriate expertise. Consideration should be given to adjunctive treatments, such as topical pro-haemostatic agents and endocrine therapy for heavy menstrual bleeding (HMB; Keeling et al, 2008; Peyvandi et al, 2013). Tranexamic acid or other anti-fibrinolytics may be sufficient for HMB and for minor bleeds, particularly at sites such as the oropharyngeal mucosa. Tranexamic acid may also be a useful adjunct to factor replacement, but is relatively contraindicated for renal tract bleeding and in cases with high thrombotic risk. For the prevention of surgical or obstetric bleeding, oral or intravenous tranexamic acid should be administered no later than 2 h before surgery or delivery to ensure peak plasma levels at the time of haemostatic challenge. Tranexamic acid should be used cautiously with prothrombin complex concentrate (PCC) or FXI concentrate because of thrombosis risk (Bolton-Maggs et al, 1994; Kohler, 1999), although more recent experience of tranexamic acid in combination with activated PCC in haemophilia inhibitor patients suggests that thrombosis risk may be low (Tran et al, 2014). Tranexamic acid is not licensed for use in children and should be used with caution in neonates. Replacement therapies for the RCD include recombinant factor concentrates of FXIII A-subunit and activated FVII (FVIIa) and plasma-derived factor concentrates of fibrinogen, FVII, FX, FXI and FXIII (Table 2). If available, specific recombinant or virally inactivated plasma-derived factor concentrates should be used in preference to fresh frozen plasma (FFP) or cryoprecipitate. Riastap (Fibrinogen) CSL Behring NovoSeven (rFVIIa) NovoNordisk Factor X (FX) Bioproducts laboratories (Elstree, UK) Factor XI concentrate (FXI) Bioproducts laboratories Hemoleven (Factor XI) LFB (Les Ulis, France) NovoThirteen (rFXIII-A) NovoNordisk Fibrogammin (FXIII) CSL Behring PCC are plasma-derived concentrates that are available as 'four factor' products containing FII, FVII, FIX and FX and as 'three factor' products without FVII (Table 3; Keeling et al, 2008). PCC may be useful in prothrombin deficiency, vitamin K-dependent clotting factor deficiency and in FX and FVII deficiency if a specific factor concentrate is unavailable. The potency of most PCC is expressed as FIX activity units, but the activities of the other constituent coagulation factors may vary between products and product batches (Table 3). High or repeated doses of PCC have been associated with arterial and venous thrombosis, usually in cases with pre-existing risk factors (Kohler, 1999). Beriplex P/N PCC (FII, VII, IX and X) CSL Behring Octaplex PCC (FII, VII, IX and X) Octapharma Standard fresh frozen plasma NHS Blood and Transplant (Watford, UK) OctaplasLG Pathogen-reduced plasma Octapharma Pathogen reduced cryoprecipitate NHS Blood and Transplant FFP is the only currently available replacement therapy for FV deficiency and combined deficiency of FV and FVIII, but may be effective in other RCD in emergencies if a more specific replacement therapy is unavailable or if diagnosis is uncertain. Cryoprecipitate may be effective in fibrinogen or FXIII deficiency if a single factor concentrate is unavailable (O'Shaughnessy et al, 2004). Should FFP or cryoprecipitate be necessary, it is currently recommended that all cases with heritable bleeding disorders receive pathogen-reduced products (Keeling et al, 2008). This requires virus inactivation either by methylene blue and light (MB-FFP) or solvent detergent (SD-FFP) treatments during manufacture, which may reduce FV, FVIII, FXI and fibrinogen content (Table 3; Williamson et al, 2003). Single donor MB-FFP supplied by UK NHS Blood and Transplant (NHS-BT) is currently quality controlled for FVIII activity (Williamson et al, 2003). The SD-FFP product Octaplas LG® (CSL Behring, Marburg, Germany) has mean FV, FVIII and FXI activities of 0·7–0·9 iu/ml and is quality controlled to ensure activities exceed 0·5 iu/ml (Table 3). The activities of other coagulation factors in SD-FFP are 0·8–1·0 iu/ml. As SD-FFP is prepared from pooled plasma donations, there is less variation in factor activities compared to single donor MB-FFP. MB-cryoprecipitate supplied by NHS-BT has an average fibrinogen content of 250 mg/unit and a minimum of 140 mg/unit. There are limited data describing the pharmacokinetics of coagulation factors administered via FFP or cryoprecipitate (Inbal et al, 1993; Horowitz & Pehta, 1998). Achieving therapeutic levels of coagulation factors, particularly with FFP, may be practically difficult because of the low starting concentration of factors in this product. Fibrinogen deficiency (F1D; MIM #202400) is an autosomal recessive or dominant disorder in which quantitative (afibrinogenaemia or hypofibrinogenaemia) or qualitative (dysfibrinogenaemia) defects in the fibrinogen Aα, Bβ or γ protein chains lead to reduced functional fibrinogen. Hypodysfibrinogenaemia describes F1D with both quantitative and qualitative fibrinogen defects. Afibrinogenaemia has an estimated prevalence of one in 1 000 000 (Mannucci et al, 2004). There are no reliable estimates of the prevalence of dysfibrinogenaemia. Fibrinogen is a complex glycoprotein comprising pairs of Aα, Bβ and γ chains and is the major ligand for the platelet αIIBβ3 integrin during platelet aggregation. Partial proteolysis of fibrinogen by thrombin enables polymerization to form fibrin clot (Weisel & Litvinov, 2013). Fibrinogen also has an anticoagulant effect, possibly by sequestering free thrombin, and contributes to fetal implantation and wound healing. Afibrinogenaemia is caused by variations in the FGA, FGB and FGG genes, which encode the fibrinogen Aα, Bβ and γ chains, respectively. Afibrinogenaemia is associated with homozygous or compound heterozygous mutations and hypofibrinogenaemia is usually linked with heterozygous mutations (de Moerloose et al, 2013). Dysfibrinogenaemia is usually associated with heterozygous mutations in FGA, FGB or FGG, clustered within specific functional domains (Haverkate & Samama, 1995; Miesbach et al, 2010; Shapiro et al, 2013). Some FGA variations cause hereditary renal amyloidosis, which is not associated with abnormal haemostasis (Gillmore et al, 2009). In 106 cases with afibrinogenaemia or hypofibrinogenaemia in US, Iranian and Indian registries, the most common symptoms were mucocutaneous, soft-tissue, joint, genitourinary, traumatic and surgical bleeding and HMB (Peyvandi & Mannucci, 1999; Acharya et al, 2004; Viswabandya et al, 2012). Intracranial bleeding was reported in 5% of registry cases. Similar symptoms and frequent umbilical bleeding were reported in 65 cases in Iranian and Palestinian case series (Fried & Kaufman, 1980; Lak et al, 1999) and in an international survey of 100 cases (Peyvandi et al, 2006). Arterial and venous thrombosis, poor wound healing and splenic rupture are rare features of afibrinogenaemia and hypofibrinogenaemia (de Moerloose et al, 2013). Some types of hypofibrinogenaemia are associated with liver disease because of retention of abnormal fibrinogen in hepatocytes (Brennan et al, 2000). In 26 cases with afibrinogenaemia or hypofibrinogenaemia in the EN-RBD registry, cases with severe bleeding had fibrinogen activity <0·9 g/l and asymptomatic cases had fibrinogen activity 0·2–2·0 g/l determined by the Clauss assay (Peyvandi et al, 2012a). Afibrinogenaemia and usually hypofibrinogenaemia manifest as prolonged prothrombin (PT), activated partial thromboplastin (APTT) and thrombin clotting (TCT) times and absent or reduced fibrinogen activity determined by the Clauss assay. There is a concordant reduction in fibrinogen antigen determined by immunoassay, gravimetric assays or by measurement of dry clot weight (Cunningham et al, 2002; Mackie et al, 2003). Assays that measure total clottable fibrinogen are an alternative that may assist diagnosis of F1D subtypes (Mackie et al, 2003). Acquired hypofibrinogenaemia is a feature of many acquired coagulopathies and can usually be distinguished from F1D on clinical grounds. Fibrinogen replacement with plasma-derived fibrinogen concentrate (Table 2) may be required to treat or prevent bleeding in F1D. In afibrinogenaemia, the recovery of fibrinogen activity after infusion of fibrinogen concentrate was 0·018 g/l per mg/kg and the half-life was 80 h, but it was shorter in children aged <16 years (Manco-Johnson et al, 2009). Therefore, a typical dose of fibrinogen concentrate of 4–6 g is expected to increase plasma fibrinogen activity by 1·0–1·5 g/l in a 70 kg adult. In a review of case reports (Bornikova et al, 2011) and in a questionnaire survey of physicians treating F1D (Peyvandi et al, 2006), fibrinogen concentrate 50–100 mg/kg every 2–4 d, to achieve fibrinogen activity >1·0–1·5 g/l was usually sufficient to treat or prevent spontaneous or surgical bleeding. Higher and more frequent dosing was required in children and in cases with severe bleeds or having major surgery (Peyvandi et al, 2006; Bornikova et al, 2011). Venous or arterial thrombosis occurred in 30% of cases in the case report series, most with afibrinogenaemia (Bornikova et al, 2011). In an open label prospective study, fibrinogen concentrate was effective in 26 bleeds and 11 surgical procedures in 12 cases with F1D. Venous thrombosis occurred in one case with other thrombotic risk factors (Kreuz et al, 2005). Fibrinogen isoantibody formation has not been reported in F1D. Pathogen-reduced cryoprecipitate has greater variation in fibrinogen content than fibrinogen concentrate and may be associated with transfusion reactions or volume overload (Table 3: O'Shaughnessy et al, 2004). A typical dose of 10–20 units (500–1000 ml) of MB-cryoprecipitate is expected to increase fibrinogen activity by 0·6–1·2 g/l in a 70 kg adult. The efficacy of cryoprecipitate is similar to that of fibrinogen concentrate (Peyvandi et al, 2006). Afibrinogenaemia may present with intracranial haemorrhage and umbilical bleeding (Lak et al, 1999; Peyvandi & Mannucci, 1999). Fibrinogen activity determined by the Clauss assay was reduced and the TCT prolonged in the healthy newborn compared to adults in some (Reverdiau-Moalic et al, 1996) but not other (Andrew et al, 1987) studies. This may be because the high sialic acid content of fetal fibrinogen affects some fibrinogen activity and TCT tests (Barr, 1978; Ignjatovic et al, 2011). Although diagnosis of afibrinogenaemia is straightforward on cord or neonatal blood samples, diagnosis of hypofibrinogenaemia requires comparison of test results with neonatal reference intervals and on re-testing at 3–6 months. Successful long-term prophylaxis with cryoprecipitate (Rodriguez et al, 1988; Peyvandi et al, 2006) or fibrinogen concentrate (Parameswaran et al, 2000; Kreuz et al, 2005; Peyvandi et al, 2006) has been reported in cases with afibrinogenaemia associated with intracranial bleeding. Typical regimens comprised fibrinogen concentrate 18–120 mg/kg, once per week to give trough fibrinogen activity of 0·5–1·0 g/l (Parameswaran et al, 2000; Peyvandi et al, 2006). Fibrinogen activity increases during normal pregnancy (Stirling et al, 1984). However, this does not prevent potential complications such as venous thrombosis, pregnancy loss, ante-partum haemorrhage (APH) and post-partum haemorrhage (PPH) in women with afibrinogenaemia and hypofibrinogenaemia (Goodwin, 1989; Dupuy et al, 2001; Roque et al, 2004; Kadir et al, 2009). Fibrinogen replacement helps maintain pregnancy and reduces bleeding. However, reports indicate that fibrinogen concentrate 5–30 g per week in 2–3 divided doses to maintain fibrinogen activity >0·6–1·0 and >1·5 g/l at delivery and post-partum does not prevent all pregnancy complications (Bornikova et al, 2011), possibly because these trough levels are inadequate. Higher doses of fibrinogen concentrate are required to maintain fibrinogen activity as pregnancy progresses (Roque et al, 2004). In a review of 250 reported cases with dysfibrinogenaemia, 53% were asymptomatic and 26% had bleeding that was typically mucocutaneous, traumatic or surgical (Haverkate & Samama, 1995). The remaining 21% of cases had venous or arterial thrombosis (Haverkate & Samama, 1995). Similar symptoms were reported in a series of 93 cases with dysfibrinogenaemia in which incidental diagnosis after routine coagulation tests or thrombophilia screens was also common (Miesbach et al, 2010; Shapiro et al, 2013). Thrombosis and bleeding may co-exist in the same case (Haverkate & Samama, 1995). Dysfibrinogenaemia may manifest as a prolonged PT and/or APTT depending on test reagent and methodology. The TCT and reptilase time are usually prolonged and there is a reduction in the Clauss fibrinogen activity, typically to 0·1–0·8 g/l. Some rare variants are associated with a shortened TCT (Cunningham et al, 2002; Mackie et al, 2003). As dysfibrinogenaemia is associated with a selective functional defect in fibrinogen activity, fibrinogen antigen or total clottable fibrinogen are not reduced. There is no association between fibrinogen activity and clinical phenotype in dysfibrinogenaemia, but some genotypes correlate with either bleeding or thrombosis (Haverkate & Samama, 1995). The PT-derived fibrinogen assay is not suitable for evaluation of dysfibrinogenaemia (Mackie et al, 2003). There are individual reports of cases with haemorrhagic dysfibrinogenaemia receiving fibrinogen concentrate (Kreuz et al, 2005). Thrombotic dysfibrinogenaemia has been managed with low molecular weight heparin for thromboprophylaxis and with coumarin anticoagulation for long-term prevention of thrombosis. Fibrinogen concentrate may have an anti-thrombotic effect by increasing the proportion of normal circulating fibrinogen molecules compared to endogenous thrombotic variant molecules. However, any therapeutic effect may be offset by the increased absolute fibrinogen concentration after concentrate infusion. There are no available data to guide optimum dosing of fibrinogen concentrate in thrombotic dysfibrinogenaemia. Women with dysfibrinogenaemia experience similar pregnancy complications to women with hypofibrinogenaemia. There are isolated case reports suggesting that these may be prevented by fibrinogen replacement throughout pregnancy (Yamanaka et al, 2003). Prothrombin (FII) deficiency (F2D; MIM #613679) is an autosomal recessive disorder in which reduced plasma prothrombin activity is caused by quantitative (hypoprothrominaemia) or qualitative (dysprothrombinaemia) defects in the FII protein. F2D has an estimated prevalence of one in 2 000 000 (Mannucci et al, 2004). FII is activated to the serine protease thrombin by activated FX in the initiation phase of coagulation, and by the prothrombinase complex in the amplification phase. Thrombin back-activates other coagulation factors and platelets and enables fibrin generation (Roberts et al, 2006). F2D is caused by variations in the F2 gene which encodes FII. There is a poor association between F2 genotype and clinical phenotype (Lancellotti et al, 2013). In 43 cases with F2D in the US, Iranian and Indian registries, the most common symptoms were mucocutaneous, soft tissue, joint and surgical bleeding and HMB. Less common symptoms were gastrointestinal, obstetric and umbilical bleeding. Intracranial haemorrhage was reported in of registry cases (Peyvandi & Mannucci, 1999; Acharya et al, 2004). Similar symptoms were reported in a survey of 26 case reports et al, 1998). Bleeding caused by F2D in was more severe in cases with FII activity iu/ml than in with FII activity iu/ml who typically bleeding (Acharya et al, 2004). In there is a poor association between clinical and laboratory et al, 2000). Heterozygous F2D carriers have FII activities of iu/ml and are typically asymptomatic et al, 1998). F2D as of the PT and APTT and reduced FII activity determined by although test results may vary by FII determined by immunoassay, is necessary to and et al, et al, 2000). assays that and may some variants et al, Acquired FII deficiency may in which is distinguished from F2D on clinical PT and APTT studies and the of et al, 2012). FII replacement with PCC may be required to treat or prevent bleeding in PCC FIX and FII activities and show FII activity recovery of iu/ml per and a half-life of h (Table 1; et al, Therefore, a typical therapeutic dose of PCC is expected to increase plasma FII activity by iu/ml. Similar doses at 2–3 intervals may be necessary for treatment 1999). If PCC is pathogen-reduced FFP is expected to increase plasma FII activity by FII have not been reported in Intracranial and umbilical bleeding may be features of F2D et al, 1999; et al, 2000). FII activity is iu/ml in healthy and at (Andrew et al, Therefore, diagnosis of F2D at delivery requires comparison of test results with neonatal reference intervals or after routine of vitamin and at re-testing There is limited experience of prophylaxis in PCC every has been reported as effective in case reports & FII activity does not during normal pregnancy (Stirling et al, and usually for delivery in women with severe pregnancy and were identified in case series of in women with F2D et al, Peyvandi & Mannucci, 1999). had comprised PCC during et al, Factor deficiency MIM is an autosomal recessive disorder in which reduced plasma FV activity is caused by quantitative qualitative defects in the FV protein. has an estimated prevalence of one in 1 000 000 (Mannucci et al, 2004). FV is in hepatocytes and in plasma and in platelet FV is activated by thrombin or activated FX to form a for activated FX in the prothrombinase complex (Roberts et al, 2006). is associated with variations in the gene that encodes FV, which usually FV & 2013). There is a poor between genotype and the clinical phenotype of In cases with in the US, Iranian and Indian registries, the most common symptoms were mucocutaneous, soft-tissue, surgical and traumatic bleeding and HMB. Less common symptoms were joint, genitourinary, and umbilical bleeding (Lak et al, Peyvandi & Mannucci, 1999; Acharya et al, 2004; Viswabandya et al, 2012). Similar symptoms were reported in series of cases et al, 2009; et al, 2011). Intracranial bleeding occurred in of registry cases and may be a feature of et al, 2000; et al, 2013). Bleeding caused by was more severe in registry cases with FV activity iu/ml than in with FV activity iu/ml who were typically asymptomatic or had mucocutaneous, surgical bleeding and HMB (Acharya et al, 2004; et al, 2009; Viswabandya et al, 2012). In 50 cases with in the EN-RBD registry, cases with severe bleeding had FV activity iu/ml and asymptomatic cases had FV activity iu/ml (Peyvandi et al, There are reports of and intracranial bleeding in cases with FV activity iu/ml (Lak et al, et al, a poor association between clinical and laboratory Heterozygous carriers have FV activity of iu/ml and are typically asymptomatic (Acharya et al, 2004). as of the PT and APTT and reduced FV activity determined by one assay. FV antigen determined by is required to qualitative from quantitative et al, 1995). FV have been reported after FFP treatment in et al, 2000; et al, and after to FV in topical thrombin in cases without & 2011). Acquired FV deficiency may be distinguished from by PT and APTT studies 1999). As there is currently no FV FV replacement with FFP may be required to treat or prevent bleeding in Pathogen-reduced FFP has been recommended for replacement therapy in although FV activity may be than FFP (Keeling et al, 2008). SD-FFP has FV activity of 0·7–0·9 iu/ml and less variation than single donor MB-FFP (Table 3). In an open label of treatment in SD-FFP increased FV activity by iu/ml and was effective for the treatment of spontaneous or traumatic bleeds and the prevention of surgical bleeds & Pehta, 1998). The half-life of FV activity after FFP infusion was h et al, & 2013). concentrates are an alternative of FV that have been used in combination with SD-FFP SD-FFP was et al, recombinant factor (rFVIIa) was effective in cases with to FFP et al, 2005; et al, with FV et al, or to volume overload et al, 2008). FV in have been using of FFP et al, or intravenous et al, transfusion was effective in acquired FV deficiency not to et al, A specific FV concentrate is currently in clinical Intracranial bleeding is reported in with usually with FV activity iu/ml et al, 2013). FV activity has a of in healthy which increases within 1 week (Andrew et al, of at delivery requires comparison of test results with neonatal reference intervals and at re-testing at months. prophylaxis with SD-FFP has been reported in with bleeding or FV activity iu/ml with FFP et al, 2000; et al, 2012). FV prophylaxis is practically difficult and may be limited by and particularly in neonates. In cases with severe it is to maintain trough FV activity using FV activity does not during normal pregnancy (Stirling et al, and is usually for delivery in women with severe was associated with in an Iranian case series (Lak et al, and in a review of in women with et al, of FFP for to maintain FV activity iu/ml and have been reported as effective in bleeding at delivery in women with et al, 2005; et al, Factor deficiency MIM is an autosomal recessive disorder in which reduced plasma FVII activity is caused by quantitative or qualitative defects in the FVII protein. has an estimated prevalence of one in 000 (Mannucci et al, 2004). of FVII as and as activated FVII which factor at sites of blood then from FVII and FX and FIX to low thrombin generation in the initiation phase of coagulation (Roberts et al, 2006). is caused by rare variations in the gene that encodes FVII (Herrmann et al, 2009). variations in the and in 30% of some and FVII activity et al, may cause clinical and laboratory in the phenotype of patients with but are to reduce FVII activity to levels associated with bleeding. In cases in the registry, the most common symptoms were mucocutaneous, soft tissue, joint and bleeding and HMB. Intracranial bleeding was reported in of cases et al, 2009; et al, 2009). Similar symptoms were reported in cases with in US, Iranian and Indian registries, although intracranial bleeding was reported in (Peyvandi & Mannucci, 1999; Acharya et al, 2004; Viswabandya et al, 2012). of cases in the and registries were asymptomatic and were identified after an abnormal PT test (Acharya et al, 2004; et al, 2009). of venous thrombosis in have et al, 2012). bleeding was more in registry cases with FVII activity iu/ml than with FVII activity iu/ml who typically had bleeding or were asymptomatic et al, 2009; Viswabandya et al, 2012). In cases with in the EN-RBD registry, cases with severe bleeding had FVII activity iu/ml and asymptomatic cases had

OBJECTIVE: Juvenile idiopathic arthritis (JIA) can persist through adolescence and adulthood, resulting in significant disability. The use of low-dose oral methotrexate (MTX) for persistent polyarthritis has been shown to be effective by the USA/USSR collaborative study group. However, 2 of the most disabling subgroups of JIA, systemic and extended oligoarthritis, were underrepresented in that study. The present study was therefore conducted to investigate the efficacy of MTX in these 2 subgroups. METHODS: Patients under the age of 16 years who fulfilled the International League of Associations for Rheumatology criteria for systemic or extended oligoarticular arthritis were eligible for this multicenter, double-blind, placebo-controlled crossover trial. Forty-five patients with systemic and 43 with extended oligoarticular arthritis were enrolled. The dosage of MTX or placebo was 15 mg/m2, which could be increased to 20 mg/m2 after 2 months. Core outcome variables were considered as primary measures, giving a final score of "improved" or "not improved." Secondary measures included scores of systemic features and biochemical laboratory measures. Assessment of function was not included since there were no validated functional measures at the start of this trial in 1991. RESULTS: In the extended oligoarticular arthritis group, MTX treatment produced significant improvement in 3 of 5 core variables (erythrocyte sedimentation rate, physician's global assessment of disease activity, and parent's global assessment of disease activity). By the primary improvement criteria, there was significant overall improvement during MTX treatment. In the systemic arthritis group, only 2 of 5 core variables were significantly improved (physician's and parent's global assessment of disease activity). Systemic features were not part of the core variables, but the systemic feature score was not significantly different between MTX and placebo treatment. There was no significant overall improvement in this group during MTX treatment. However, no significant interaction between disease subgroup and treatment effect was demonstrated. When the data from both disease subgroups were combined, there was significant clinical improvement during MTX treatment (P = 0.006). CONCLUSION: MTX 15-20 mg/m2 given orally once a week was found to be an effective treatment for both extended oligoarticular and systemic JIA in this shortterm trial. Long-term efficacy needs to be addressed in future studies.

The objective of this guideline is to provide healthcare professionals with guidance on the management of patients with primary autoimmune haemolytic anaemia (AIHA). The guidance may not be appropriate to every patient and in all cases individual patient circumstances may dictate an alternative approach. Attempts to categorise autoimmune haemolytic anaemia (AIHA) and define its response to treatment vary considerably in the published literature. Author defined criteria have been used in this guideline, but this limits study comparisons and will have contributed to differences in reported outcome. The investigation and diagnosis of adult and paediatric AIHA are considered together. Guidance on the treatment of adult AIHA is then followed by a section on paediatric AIHA. Recommendations are based on the systematic review of published English language literature from January 1960 to October 2015 (see Appendix S1 for further details). Although recommendations are unchanged, an expanded version of this guideline is available as Appendix S2. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) nomenclature was used to evaluate levels of evidence and to assess the strength of recommendations. The GRADE criteria are specified in the British Committee for Standards in Haematology (BCSH) guidance pack (http://www.bcshguidances.com/BCSH_PROCESS/42_EVIDENCE_LEVELS_AND_GRADES_OF_RECOMMENDATION.html) and the GRADE working group website http://www.gradeworkinggroup.org The guideline group was selected to be representative of UK-based experts in the diagnosis and management of AIHA. Review of the manuscript was performed by the BCSH General Haematology Task Force, BCSH Executive Committee and then a sounding board of the British Society for Haematology (BSH). This compromises 50 or more members of the BSH who have reviewed this Guidance and commented on its content and applicability in the UK setting. AIHA is a decompensated acquired haemolysis caused by the host's immune system acting against its own red cell antigens. The incidence is approximately 1 per 100 000/year (Pirofsky, 1975; Klein et al, 2010). It can occur at any age but incidence rises with increasing age. Serologically, cases are divided into warm type (65%), cold type (29% cold haemagglutinin disease [CHAD], 1% paroxysmal cold haemoglobinuria) or mixed AIHA (5%). Approximately half are primary (idiopathic) AIHA and half are secondary to associated disorders (Table 1). Patients with AIHA may present with symptoms of anaemia (weakness 88%, dizziness 50%, dyspnoea 9%), haemolysis (jaundice 21%, dark urine 3%) or symptoms of an underlying disorder (Pirofsky, 1975). Without underlying disease, examination may be unremarkable or reveal mild pallor or splenomegaly. Less often, severe haemolysis leads to hepatosplenomegaly, haemoglobinuria and signs of heart failure (Packman, 2008). Cold haemagglutinin disease (CHAD) can present as a primary chronic clonal disorder, usually occurring in middle age or in the elderly. Cold-induced acrocyanosis (dusky blue appearance of toes, fingers, nose tip or ears) or Raynaud phenomenon occur in 40‒90% of patients (Berentsen et al, 2006; Swiecicki et al, 2013). Secondary CHAD can be self-limiting, for example following childhood infection. With its different natural history, secondary CHAD has also been termed cold agglutinin syndrome (Berentsen & Tjonnfjord, 2012). Paroxysmal cold haemoglobinuria (PCH) is typically transient, presenting 1–2 weeks after an upper respiratory tract infection or other childhood illness with acute fever, abdominal, back or leg pain and haemoglobinuria (Gehrs & Friedberg, 2002). Haemolysis can be severe and intravascular but usually settles over several weeks. When a patient presents with suspected AIHA, three questions should be considered. Is there haemolysis; is the haemolysis autoimmune and what is the type of AIHA? However, there may be confounding factors as these laboratory tests are not highly specific. Some parameters may be normal, especially with mild compensated haemolysis. The differential diagnosis of haemolytic anaemia is shown in Table 2. A positive direct antiglobulin test (DAT) indicates the presence of immunoglobulin (Ig)G, IgM, IgA or complement (usually C3d) bound to the red cell membrane. In the presence of haemolysis, this suggests an immune aetiology but clinical assessment is required before a diagnosis of AIHA can be made. Typically monospecific anti-IgG and anti-C3d antibodies are used in the initial screening and these help to determine the type of AIHA. A positive DAT is not specific and is also associated with a wide range of non-haemolytic disease states, possibly through passive deposition of immunoglobulins or immune complexes; examples include liver disease, chronic infection, malignancy, systemic lupus erythematosus (SLE), renal disorders and drugs such as intravenous immunoglobulin (IVIg) or antithymocyte globulin. Rarely, AIHA patients test negative with a tube test DAT, for example due to a low affinity antibody, low levels of red cell bound antibody or an immunoglobulin not tested for (e.g. IgA-only AIHA). A gel column agglutination method is a more sensitive method that is less prone to error than a conventional tube test (Fayek et al, 2012). AIHA can be diagnosed in 3% of patients testing negative with a gel card method by using a red cell elution technique (Sachs et al, 2006). The Donath-Landsteiner test may be considered in children with haematuria and is discussed under investigations. Patients with DAT-negative AIHA generally have a milder anaemia and are steroid responsive. The most relevant tests to investigate for an underlying cause for AIHA are shown in Table 3. Although reticulocytopenia can occur in the acute phase of AIHA; haematinic deficiency, marrow infiltration, aplastic anaemia and parvovirus B19 infection should be considered if it is present. Further serological investigation is required to determine the type of AIHA (e.g. warm, CHAD, PCH) as the approach to treatment differs. Finally, if the patient requires blood, investigations are needed to exclude underlying alloantibodies and identify units suitable for transfusion. In adults, two 7 ml EDTA samples are usually sufficient for initial serological investigation. A clotted sample is also required for investigation of suspected PCH or DIIHA. If DAT is positive for C3 ± IgG and i)DAggT negative or insignificant CAs and ii)age <18 years or haemoglobinuria or cold associated symptoms or atypical serology Typical serological characteristics of AIHA subtypes are shown in Table 4. Although the autoantibody specificity can sometimes be identified, specificity does not help predict the clinical outcome (Issitt, 1985). Is caused by autoantibodies (usually IgG) that bind red cells optimally in vitro at 37°C. When tested with anti-C3 and anti-IgG reagents, the DAT would be positive for: IgG only (35%), IgG + C3 (56%) or C3 only (9%) (Issitt, 1985). AIHA can be considered warm when there is a consistent clinical picture and a DAT positive to IgG, C3 or both, when a clinically significant cold reactive antibody has been excluded (Fig 1). Is caused by autoantibodies (usually IgM) that bind red cells optimally in vitro at 4°C. Although the DAT is usually positive for C3 only, 21–28% are also positive with IgG (Berentsen et al, 2006; Swiecicki et al, 2013). Furthermore, only 7–31% of patients with AIHA and a C3-only positive DAT have CHAD. Marked red cell agglutination on the blood film is classically seen in CHAD but can occur in mixed AIHA and PCH. Milder agglutination sometimes occurs in warm AIHA and clinically insignificant polyclonal cold agglutinins (CAs) can cause agglutination on a blood film spread at room temperature. Up to 35% of patients with warm AIHA have CAs reactive at 20°C (Petz & Garratty, 1980). CHAD must therefore be distinguished from insignificant CAs. The thermal amplitude of CAs (the maximum temperature at which antibody binds red cells in vitro) is usually <25°C. At 4°C, the CA antibody titre is usually only positive with a dilution <1:64 and it rarely exceeds 1:256. In CHAD, the titre is usually >1:500 at 4°C and the thermal amplitude ≥30°C (but can be as low as 25°C if red cells are suspended in saline rather than 30% bovine albumin). Defining an absolute cut-off for titre or thermal amplitude is difficult and there are exceptions. CHAD can be diagnosed in patients with AIHA and a DAT positive to C3 ± IgG, with a consistent clinical picture and a high titre cold reactive antibody. The thermal amplitude may be considered as a supportive serological investigation where diagnostic uncertainty exists. The term ‘primary’ CHAD has been used to describe patients without other systemic autoimmune disease or infective aetiology and who have no clinical or radiological evidence of underlying lymphoma. However, with immunophenotyping, the majority of such cases have evidence of a clonal bone marrow lymphoproliferative disorder and a circulating IgM monoclonal paraprotein (Berentsen, 2011). The paraprotein can be detected by serum electrophoresis and immunofixation in >90% of cases (Berentsen et al, 2006) but the sample must be kept at 37°C until the serum has been separated or the antibody will remain bound to red cells. All cases of suspected primary CHAD should be reviewed by an appropriately constituted haemato-oncology multidisciplinary team (National Institute for Clinical Excellence, 2003). PCH is caused by a biphasic IgG antibody that binds to red cells at low temperature and causes complement-mediated lysis as the temperature is raised. The DAT is usually positive to C3 only. There may be agglutination, spherocytes or erythrophagocytosis by neutrophils on the blood film. Reticulocytopenia is common early in PCH, evolving into reticulocytosis with recovery. PCH can be diagnosed in patients with AIHA and a positive Donath-Landsteiner test. The test can be technically difficult (Sokol et al, 1999) and false negative results can be avoided by using an indirect method. Testing should be performed by a specialist laboratory and a warm separated serum sample is required. Testing should be considered in patients with AIHA and a DAT positive for C3 ± IgG, when CHAD has been excluded, and there is either haemoglobinuria, cold-associated symptoms, atypical serological features or if the patient is <18 years old. The DAT is negative in some cases of PCH. The Donath-Landsteiner test should therefore also be considered in children with haemolysis, haemoglobinuria and a negative DAT. Mixed AIHA is caused by a combination of a warm IgG antibody and a cold IgM antibody with a thermal amplitude of at least 30°C. The DAT is usually positive with IgG and C3. The cold antibody may have a low antibody titre (e.g. <1:64). Cold-induced haemolysis, Raynaud phenomenon or acrocyanosis do not appear to be features of mixed AIHA (Sokol et al, 1983; Shulman et al, 1985). Mixed AIHA can be diagnosed in patients with AIHA, a DAT positive for IgG and C3, a cold antibody with a thermal amplitude ≥30°C, evidence of a warm IgG antibody and the absence of typical features of CHAD. A diagnostic pathway is illustrated in Fig 1. Patients with AIHA and a DAT positive for C3 ± IgG should be screened for a cold antibody. A direct agglutination test (DAggT) can be performed as a screening test in the local transfusion laboratory; a clinically significant cold haemagglutinin can be excluded if saline-suspended normal red cells are not agglutinated by the patient's serum after incubation at room temperature for 2008). If this screening is further testing is needed to insignificant CAs from CHAD. for and thermal amplitude should be kept at 37°C for this can be samples should be to 37°C in a for 1 before testing (Issitt, 1985). The diagnostic (Fig is a and the diagnosis is not The clinical picture should be considered and the of a laboratory may be required before a diagnosis is made. A of serological testing antibody thermal amplitude and the Donath-Landsteiner is the absence of a UK Testing should therefore be in these tests on a should be by and of BCSH on et al, 2013). The of the investigation are to determine and of the patient and identify if present. testing can or more Approximately 30% of patients with AIHA have an underlying or but these are if there is no of transfusion or If anaemia is transfusion with and blood is more appropriate than until serological investigations have been In patients with a clinically significant cold type antibody, the of a blood and a warm for transfusion is the evidence of is are the available in an from suggests that of patients to for and most for weeks et al, was by low and is in the UK of as a term treatment when the is (but in patients with or as a to et al, 2011). The evidence for is to and any is has been used in patients with severe haemolysis with other such as et al, 2010). The of high intravenous is to may have a in cases but the of may also et al, 2006; et al, Patients with severe haemolysis who have not to may If the patient is not weeks to this should be until as antibody are et al, 2011). In patients with warm AIHA for (e.g. of anaemia or of available blood to have with are the available in an The response of CHAD to can be with response of in are and be without an high steroid However, a of 1 may be considered as a seen in 2003). However, are and warm AIHA, its may be in patients with severe disease in with alternative can occur the cell and its especially if the agglutinin is at 37°C and the room and may a high temperature setting. or alternative of with has been et al, 2010). is an cause of and in AIHA and is more when haemolysis is In study of patients with severe AIHA as in and was more if no was 2003). In in haemolysis range and et al, to the of cases of anaemia and reported in patients with chronic haemolytic The incidence of disease in the is approximately with of upper increasing to with factors include increasing and In patients the is in drugs or of may also occurs in to of term et al, 2012). and and units bone and are for all patients et al, 2012). In bone was by and years with an at a of are considered high for and treatment such as a is et al, et al, et al, 2012). are in Fig 2. AIHA is a chronic and the of is disease with Patients with mild compensated haemolysis may not and are but from haemolysis can the of infection in patients on is Approximately of patients to at a to and approximately The initial response may several weeks but absence of response by should be considered a steroid In an can for example or after a maximum of to over and then by every In a of primary AIHA was more common if to in less than and if in less than et al, 2010). Approximately of patients remain in after are Although a further can an on due to the term of should be considered. are but do not that is to et al, et al, 2010). The and most used are and Approximately of cases to but response are reported with or patients and the of infection et al, et al, et al, the of infection and chronic of AIHA, most patients will from an to of of have been reported following for primary warm AIHA et al, et al, et al, In primary and secondary warm AIHA, with in et al, does not outcome is associated with of AIHA. In the only and was to et al, 2013). At to response is approximately weeks The term is but occurs in after a of et al, et al, and in by et al, 2013). is severe et al, or have been of is a and screening with serology for and antibody is 1 is a et al, The is for weeks but low cell when used for autoimmune disorders et al, 100 for weeks with or et al, response However, was used at an disease than of and of response and further The treatment are as to no for a Approximately of AIHA patients to et al, with (Pirofsky, or et al, et al, However, the steroid and the of response is and should be excluded to and some in AIHA. the was typically of 7 patients with primary warm AIHA to to et al, 1985). The was expanded to with a response In a further patients with secondary AIHA to three and some in primary and secondary AIHA. patients and with to 1 typically of patients with AIHA of patients or to appear in primary secondary AIHA. If the are of primary warm AIHA cases to & & & & et al, 1999) with a of occur the of but have been Approximately a of patients after Although early that high would predict a response to this was not by and has from clinical and the should be based on the of or should be and a of for at et al, 2011). Approximately of patients of with in with haemolytic anaemia et al, 2010). or is also more common in with haemolytic occurring in of et al, low has been on of et al, but evidence of is term is also by AIHA and some in AIHA have (see Appendix Although some has been reported with low (e.g. with or without there are on or its be over intravenous also appear for example 50 for et al, or 1 for & that and is should be to selected patients with disease following multidisciplinary than patients for AIHA have been reported to the for and Some have been with & 2008). should be performed in Committee of and with in for patients with autoimmune AIHA caused by IgA occurs in of cases and usually to conventional treatment and Mixed AIHA is usually as severe haemolysis et al, Approximately are primary secondary cases are associated with Mixed AIHA is steroid but most leads to chronic haemolysis. was in (Sokol et al, and patients et al, 1985). has been reported with and for acute haemolysis, with for underlying and with all cold where to the of severe to the and in cold should be considered for severe symptoms or transfusion (Berentsen & Tjonnfjord, 2012). In patients has usually been avoided IgM red cells are not in the of is therefore and to have a CHAD is less than warm AIHA. or do not the of or was in some but not all also a response to et al, et al, and et al, 2013). In the response to for weeks was (Berentsen et al, (Berentsen et al, and secondary et al, and treatment was with a response of In a study of with the response was and response (Berentsen et al, 2010). such as and can haemolysis in CHAD can by of temperature. the antibodies thermal amplitude may help define a temperature should be if a CHAD is on may and of the heart to insignificant CAs then In patients with CHAD or CAs a of such as warm with systemic have been can also present with in the et al, et al, or with agglutination in the However, appear in patients with CAs without to et al, and serological screening to cold is by it is not AIHA can occur at any age childhood from through to but with a incidence In to of cases it is a only term et al, AIHA in children followed in by PCH, typically by a infection. CHAD is less common in children to adults, and a infection. disease syndrome or is associated with approximately of children present with or dark Less there will be or pain and 3% with or acute renal due to severe anaemia et al, 2011). The laboratory investigations and differential diagnosis are in the adult section and in and 3. In the differential should be to such as of childhood and parvovirus B19 infection. lymphoproliferative syndrome and a primary should be tested for before or should include serum cell and is cell and liver tests should also be et al, 2011). The management of AIHA in children is to that in the adult is typically as at a of et al, & 1983; et al, with in of children with primary or secondary AIHA. children for a blood transfusion can be less until there is evidence of which is when the is & may be a in children to of for et al, The is with response of in children with primary or secondary AIHA. A response to was reported in et al, (Sokol et al, and et al, However, there was on of response in of but three of a of & that childhood AIHA is self-limiting, should usually be considered a treatment and that and may also have some in childhood AIHA. In a children to as treatment for primary AIHA et al, The most common of PCH is acute and transient, following an illness in PCH may for to of AIHA in Although the cases of PCH in patients with or chronic cases are usually or infection. cases in children have a of a upper respiratory infection. include and Clinical features are and diagnostic tests are in the section on investigation of AIHA. to the of PCH, initial management is In the acute intravascular haemolysis can be severe and blood transfusion may be required. blood is not usually required (Sokol et al, should be with but should be avoided due to the of haemolysis. Although cold has been there is no evidence to the of Patients can a without which are for severe or In the of disease, may the haemolysis. AIHA has been to occur in approximately 1 in (Sokol et al, and investigation for underlying causes should be to cases but usually avoided and causes of haemolysis especially if there is a outcome is generally and cases or after IgG autoantibodies can the and cause or haemolysis to haemolytic disease of the In the there or at & Although the majority of have no following the DAT is positive and antibodies may in anaemia and in the or anaemia weeks. patients typically to high may have an on the There is evidence to treatment of Some patients with CHAD with treatment (e.g. warm, and blood Some considered in such as and have been used in AIHA can the in a of who an all with no or et al, 2011). for anaemia can be by of the middle et al, In the of transfusion is to a of anaemia blood is to be negative for the autoantibody and of haemolysis and any underlying cause should be to haemolysis in et al, and may have a in AIHA with to the circulating autoantibody and early also be but are with early anaemia and have been with but sometimes these cases required transfusion. In early of the for transfusion & 2003). anaemia in the mild and in cases of AIHA. In to the BSH and the BSH group reviewed and commented on the and All in and of the All the version of the The would to the BCSH the BSH sounding board and the BCSH for in these All have a of to the BCSH and Task which may be reviewed on The following members of the group have no of to and of the group will the group if any evidence available that would the strength of the recommendations in this or it The will be and from the BCSH website if it If recommendations are an will be published on the BCSH website at If are required due to in of evidence or significant evidence recommendations a version of the guidance will be on the BCSH the and in these is to be and at the of to the the British Committee for Standards in Haematology (BCSH) the any for the content of these Appendix review for the guideline on diagnosis and management of primary autoimmune haemolytic Appendix S2. version of the BSH guideline on the diagnosis and management of primary autoimmune haemolytic The is not for the content or of any by the than should be to the for the