Bleeding symptoms and coagulation ... - Wiley Online Library

0 downloads 129 Views 223KB Size Report
FX:Ag/FX:C 2´5, range 0´81±9´25, n ˆ 16) than cases with normal FX:C .... (Dade), actin FS (Dade) and 6´6 U/ml bov
British Journal of Haematology, 2000, 110, 454±460

Bleeding symptoms and coagulation abnormalities in 337 patients with AL-amyloidosis Andrew D. Mumford, 1 James O'Donnell, 1 Julian D. Gillmore, 2 Richard A. Manning, 1 Philip N. Hawkins 2 and Michael Laffan 1 1 Department of Haematology and 2 Immunological Medicine Unit, Division of Medicine, Imperial College School of Medicine, Hammersmith Hospital, London, UK Received 25 February 2000; accepted for publication 10 April 2000

Summary. Haemorrhage is a frequent manifestation of amyloidosis. We performed a retrospective clinical analysis of 337 patients with systemic immunoglobulin light-chain (AL)-amyloidosis, in whom whole-body serum amyloid P component (SAP) scintigraphy and a clotting screen had been performed. Abnormal bleeding was noted in 94 cases (28%), and the coagulation screen was abnormal in 172 cases (51%). The most common abnormalities were prolongation of the thrombin time (TT; 108 cases, 32%) and the prothrombin time (PT; 82 cases, 24%). In multivariate analysis, a prolonged PT was the only coagulation abnormality associated with abnormal bleeding (P ˆ 0´0012), but this was independent of the whole-body amyloid load. Prolongation of the TT was associated with hepatic amyloid infiltration (P , 0´00001), with proteinuria (P , 0´001) and low serum albumin (P , 0´00001). In 154 patients who were studied further, subnormal factor X activity (FX:C) was found in 22 cases (14%). In cases with

subnormal FX:C, the corresponding factor X antigen (FX:Ag) measurements were consistently higher (median FX:Ag/FX:C 2´5, range 0´81±9´25, n ˆ 16) than cases with normal FX:C (median FX:Ag/FX:C 0´96, range 0´65± 1´29, n ˆ 28, P , 0´0001). No evidence was found of an FX inhibitor. Of the 48/154 (31%) cases with a prolonged TT, the reptilase time was also prolonged in 38/48 cases (79%). These data show that haemorrhage and abnormal coagulation are common in AL-amyloidosis and are multifactorial in origin. We provide evidence suggesting that hepatic amyloid infiltration and nephrotic syndrome are determinants of the TT. In most patients, prolongation of the PT was explained by reduction in FX:C, but this was not wholly explained by a reduction in FX:Ag.

Abnormal bleeding is frequently encountered in the management of patients with immunoglobulin light-chain (AL)-amyloidosis and, although mild intracutaneous haemorrhage is most common, life-threatening bleeding is well described (Kyle & Greipp, 1983; Yood et al, 1983). Whereas all patients with systemic amyloidosis may have widespread small vessel fragility as a direct result of amyloid infiltration (Greipp et al, 1981), the risk of bleeding in patients with ALamyloidosis can be further exacerbated by a variety of acquired abnormalities of coagulation. Deficiencies in specific coagulation factors in AL-amyloidosis have long been recognized (Korsan-Bengtsen et al, 1962) and, although factor X deficiency has been studied most widely, there are also anecdotal reports of acquired deficiencies in factor IX (Korsan-Bengsen et al, 1962; McPherson et al, 1977;

Greipp et al, 1981), factor II (McPherson et al, 1977; Greipp et al, 1981), factor VII (McPherson et al, 1977) and factor V (Gatel, 1998). Multiple factor deficiencies and hypofibrinogenaemia have been described in amyloidosis in association with disseminated intravascular coagulation and increased fibrinolysis (Perlin et al, 1971; Liebman et al, 1983). Abnormalities of fibrin polymerization are common in the lymphoproliferative disorders, and one study reported that a thrombin inhibitor was the most frequent coagulation abnormality in AL-amyloidosis (Gastineau et al, 1991). Evidence that clotting factors may bind to amyloid (Furie et al, 1977, 1981) and anecdotal reports of improved haemostasis after the removal of severely amyloidotic spleens (Greipp et al, 1981) have raised the possibility that clotting factors may be sequestered within amyloid in vivo. Suitable methods have not existed previously for testing the hypothesis that the severity of haemostatic defects is thus related to the whole-body amyloid load. We have investigated the relationships between the

Correspondence: Dr M. Laffan, Department of Haematology, Imperial College School of Medicine, Hammersmith Hospital, Du Cane Road, London W12 ONN, UK. E-mail: [email protected].

454

Keywords: AL-amyloidosis, serum amyloid P scintigraphy, factor X deficiency, thrombin time, fibrin polymerization.

q 2000 Blackwell Science Ltd

Haemostatic Disorders in AL-amyloidosis complex haemostatic defects in amyloidosis by performing a systematic analysis of clinical and coagulation data from a large series of amyloid patients, uniquely including a measurement of whole-body amyloid load using serum amyloid P component (SAP) scintigraphy. We report that abnormalities in coagulation are common and that, in most patients, these can be explained by either impaired fibrin polymerization or reduction in factor X activity and are not directly related to the extent of amyloid deposition. MATERIALS AND METHODS Patients. In patients with suspected amyloidosis referred to the Immunological Medicine Unit at Hammersmith Hospital, the diagnosis of AL-amyloidosis was established by a combination of histology, with immunohistochemical exclusion of amyloid A protein (AA) amyloidosis and transthyretin-related amyloid, compatible clinical findings, the absence of a family history of amyloid and a demonstrable monoclonal B-cell dyscrasia. In approximately 10% of patients, a B-cell clone was not identified per se. All patients underwent full clinical assessment including specific enquiry about bleeding symptoms and their timing with respect to the onset of AL-amyloidosis. Patients were phlebotomized for basic haematological and biochemical investigations, and a 24-h urine collection was performed for estimation of protein concentration. The presence of a paraprotein was sought and, if possible, quantified by electrophoresis and immunofixation of serum and urine. All subjects were tested for prothrombin time (PT), activated partial thromboplastin time (aPTT) and thrombin time (TT). A semiquantitative assessment of total amyloid load and its anatomical distribution was made for each patient with SAP scintigraphy: visual scoring of whole-body amyloid load was graded as: `small' when the tracer signal was discernibly above normal blood pool levels in amyloidotic organs but could be displayed within the usual blood pool grey-scale; `medium' when the intensity of abnormal uptake was substantially greater than the blood pool signal but the latter could still be visualized when the grey-scale was adjusted to encompass target organ uptake; and `large' when the abnormal signal was so intense that definition of blood pool was lost when the grey-scale was adjusted to encompass target organ uptake (Rydh et al, 1998). Between June 1990 and December 1998, 351 patients received a diagnosis of AL-amyloidosis and, of these, 337 patients were evaluable for clinical history, SAP scan and clotting screen (PT, aPTT and TT). Six patients receiving oral anticoagulants at the time of the outpatient consultation were excluded, as were those with haemostatic abnormalities that clearly anteceded the development of amyloid (n ˆ 2). Six further patients were excluded because of failure to obtain sufficient blood for the clotting screen at the first outpatient consultation. A further study group of 154 patients who presented sequentially after May 1997 were investigated in greater detail by performing specific factor assays as detailed below. Specimen collection. Venous blood was collected directly into 0´105 mol/l trisodium citrate and centrifuged at 48C

455

for 10 min at 3000 r.p.m. An aliquot of platelet-poor plasma was immediately separated for PT, aPTT and TT, and the remaining plasma was stored in aliquots at 2808C for further analysis where appropriate. Coagulation assays. Before January 1997, samples were analysed using an ACL300R (Instrumentation Laboratory) and KC10 (Brownes) using calcium rabbit brain thromboplastin (Diagnostic Reagents), kaolin/platelet substitute mixture (Diagnostic Reagents) and 6´6 U/ml bovine thrombin (Diagnostic Reagents) for PT, aPTT and TT respectively. After January 1997, samples were analysed using a Sysmex CA6000 using Innovin recombinant thromboplastin (Dade), actin FS (Dade) and 6´6 U/ml bovine thrombin (Diagnostic Reagents). Factor assays, fibrinogen and reptilase time (RT). All samples from the further study group underwent factor V and X activity assays (FV:C and FX:C respectively) by a one-stage method with factor V- and X-deficient substrate using the CA6000. The following reagents were used: factor V- and Xdeficient plasma (Immuno), 1´0 iu/ml reference standard (Immuno) calibrated against WHO international standard, normal control plasma (Immuno), abnormal control plasma (Alpha Laboratories) and Innovin recombinant thromboplastin (Dade). All patients with FX:C , 0´50 iu/ml underwent repeat testing at 1:2 and 1:4 dilutions in FX-deficient plasma. Clauss fibrinogen assays were performed on the CA6000 using 100 U/ml bovine thrombin (Diagnostica Reagents). The RT (Payne-Burns) was measured in all those samples with prolongation of the TT, using the same operating conditions as the TT. Factor X antigen assay (FX:Ag) was performed on 44 samples, including all those from individuals with subnormal FX:C and from a random selection of those with normal FX:C using a rabbit antifactor X antibody enzyme-linked immunosorbent assay (ELISA; Asserchrom X:Ag Diagnostica Stago). Normal ranges and statistical analysis. A clotting time was considered abnormal if it lay above the laboratory normal ranges (before January 1997: PT, 12´0±15´5 s; aPTT, 30± 46 s; TT, 15±19 s; after January 1997: PT, 9´6±11´6 s; aPTT, 24±32 s; TT, 15±19 s). Because of the different normal ranges within the study group, clotting times are henceforth expressed as a ratio with the normal range median as the denominator. The PT and aPTT ratios from approximately the first 100 patients in the study, which were expressed relative to the pre-1997 normal ranges, were adjusted to equivalent post-1997 times by a simple transformation based on the ratio of the standard deviations of the pre- and post-1997 normal range populations. Equivalent measurements of PT, aPTT and TT were then compared by univariate analysis with the following clinical and laboratory findings: bleeding history, whole-body amyloid load assessed by SAP scintigraphy, splenic and hepatic amyloid deposition by SAP scintigraphy, serum creatinine, serum albumin, total 24-h urine protein, presence of a serum or urine paraprotein and, if present, paraprotein concentration and subclass. Variables that showed significant associations were re-examined by multiple linear regression to establish independent measures of

q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 454±460

456

A. D. Mumford et al

Table I. Laboratory characteristics of the study group (n = 337). Biochemical parameters

Median

Range

Normal range

Serum creatinine (mmol/l) Serum albumin (g/l) 24 h urine protein (g) Platelet count (  109/l) Fibrinogen (g/l)

101 31 2´1 312 3´9

43±1053 10±49 0±27 15±512 0´4±8´6

55±125 35±55 , 0´2 150±400 2±4

Biochemical parameter Serum paraprotein None detected IgG IgA IgM Light chain only SAP scintigraphy Whole body load Anatomical site

n (%)

Large Medium Small Splenic amyloid Hepatic amyloid

192 98 24 10 13

(57) (29) (7) (3) (4)

115 64 158 236 165

(34) (19) (47) (70) (49)

Fig 1. Proportions of study group with prolongation of PT, aPTT and TT. Total proportions of patients with at least one prolonged clotting time are shown boxed, and proportions with different combinations of prolonged PT, aPTT or TT are shown unboxed.

association. There were no differences in the frequencies of abnormal coagulation results between the pre- and postJanuary 1997 patients. RESULTS The study group of 337 patients comprised 182 men (54%) with a median age at diagnosis of 61´2 years (range 28´9±86´5 years). The clinical and laboratory characteristics of the group are shown in Table I. Nephrotic syndrome (urine protein . 3 g/24 h) was present in 94 patients (28%). The number of patients whose clinical records indicated episodes of bleeding after the development of amyloidosis is shown in Table II. Bleeding was the presenting feature of Table II. Frequency of bleeding symptoms in the study group. Bleeding site

n (%)

No bleeding Cutaneous bleeding Gastrointestinal Renal tract Post-procedure (dental  2, laryngoscopy, arthroscopy, renal biopsy) Retroperitoneal Haemoptysis Epistaxis Vaginal Intrasplenic Intrahepatic

243 60 16 5 5 2 2 1 1 1 1

(72) (18) (5) (1) (1)

AL-amyloidosis in 10 patients (five gastrointestinal, four cutaneous and one intrasplenic). Three patients presented after the incidental discovery of an abnormal coagulation screen. Prolongation of the PT, aPTT or TT was observed in 172/ 337 (51%) patients. In 60/337 (18%) patients, there was prolongation of more than one clotting time. The two most common abnormalities were prolongation of the TT, observed in 108/337 (32%) patients, and prolongation of the PT in 82/337 (24%) patients. Although prolongation of the aPTT was observed in 46/337 (14%) patients, there was co-existent prolongation of the PT in 34/46 (74%) patients (Fig 1). Thrombocytopenia (platelet count , 150  109/l) was present in 10/337 (3%) patients (Table I). Six thrombocytopenic patients had myeloma and four patients had thrombocytopenia of undetermined origin. In the multiple linear regression analysis, prolongation of the TT was associated with the presence of hepatic amyloid deposits as assessed by SAP scintigraphy (P , 0´00001), with increased 24-h urine protein (P , 0´001) and with low serum albumin (P , 0´00001). Prolongation of the PT was associated with a history of abnormal bleeding at sites other than the skin (P ˆ 0´0012). There were no further independent associations among the PT, aPTT or TT, the presence of amyloid deposits at other anatomical sites, the presence or subtype of paraprotein or any other laboratory markers of kidney or liver function. There was no association between whole-body amyloid load and prolongation of the PT or aPTT. Although on univariate analysis large whole-body amyloid load was significantly associated with prolongation of the TT, the association with hepatic amyloid was more significant in the multiple linear regression, and so the second of these non-independent variables is likely to be the causal association. There was no association between thrombocytopenia and bleeding or any

q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 454±460

Haemostatic Disorders in AL-amyloidosis

457

Fig 4. Distribution of fibrinogen concentrations in the 154 patients in the further study group according to TT showing upper and lower limits of the normal range. Fig 2. Distribution of FX:C and FV:C showing upper and lower limits of the normal ranges.

of the measured laboratory parameters including the presence of splenic amyloid. The median FX:C in the 154 patients studied further was 1´0 iu/ml (range 0´04±1´86 iu/ml). In 22/154 (14%) patients, FX:C was less than 0´7 iu/ml, and in 7/154 (5%) patients FX:C was less than 0´20 iu/ml (Fig 2). Fifteen out

Fig 3. Relationship between FX:C and FX:Ag in 44 patients from the further study group. The solid lines correspond to FX:Ag/FX:C ratios of 1´25 and 0´75. The majority of patients with low FX:C had FX:Ag/FX:C ratios . 1´25.

of 22 (68%) patients with FX:C , 0´7 iu/ml had splenic amyloid demonstrated on SAP scan compared with 95/132 (72%) of those with higher FX:C. The median FV:C was 1´35 iu/ml (range 0´36±2´24 iu/ml), with only 3/154 (2%) patients below 0´6 iu/ml. There was no difference in the frequency of abnormal coagulation results or in the clinical or laboratory characteristics of the further study group and the group as a whole. All individuals with subnormal FX:C in the further study group showed prolongation of the PT with or without prolongation of the aPPT. Two patients showed prolongation of the PT with normal FX:C and factor VII activities of 0´44 and 0´23 iu/ml (normal range 0´6±1´5 iu/ml). A further 8/154 (5%) individuals showed normal FX:C and a prolonged PT for which no cause could be established. In all these patients, the prolongation of the PT was less than 3 s. Similarly, one individual from the further study group had a prolonged aPTT and normal FX:C, and, in this case, factor XII activity was 0´45 iu/ml (normal range 0´5±1´5 iu/ml). Within the subgroup of patients with subnormal FX:C, there was no demonstrable relationship between FX:C and the magnitude of the prolongation of the PT or the clinical severity of bleeding episodes. In 15 individuals with subnormal FX:C, the FX:C assay was repeated at different dilutions with FX-deficient plasma to detect the presence of an inhibitor. In 15/15 (100%) cases, the result from each dilution corresponded within the CV (coefficient of variation) of the assay, indicating that no inhibitory activity was present. The relationship between FX:C and FX:Ag in the 44 tested patients is shown in Fig 3. For patients with FX:C within the normal range (n ˆ 28), there was a strong correlation between FX:C and FX:Ag (r ˆ 0´82, median FX:Ag/FX:C ratio 0´96, range

q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 454±460

458

A. D. Mumford et al

Fig 5. Relationship between TT and RT in the 48 patients from the further study group with prolongation of TT. Dashed lines indicate the upper limit of the normal range.

0´65±1´29), but of those patients with FX:C , 0´7 iu/ml (n ˆ 16) 15/16 (94%) had higher FX:Ag (median FX:Ag/ FX:C ratio 2´5, range 0´81±9´25, P , 0´0001). Within the further study group, 48/154 (31%) patients had a prolonged TT and, in these patients, the fibrinogen Clauss assay results were significantly higher (median 4´83 g/l, range 3´02±6´43 g/l) than in patients with a normal TT (median 3´84 g/l, range 1´74±5´46 g/l, P , 0´0001, Fig 4). Thirty-eight out of 48 (79%) patients with a prolonged TT showed prolongation of the RT but, in 10/48 (21%) patients, the RT was normal. The relationship between TT and RT is shown in Fig 5. DISCUSSION The clinical features of AL-amyloidosis are protean and, without treatment, the clinical course is dominated by progressive organ failure resulting from a variable pattern of amyloid deposition in the heart, spleen, liver, kidneys, gut, nerves and other tissues. The laboratory characteristics of our study group reflect this clinical diversity and are similar to those of a previously reported survey of patients with ALamyloidosis (Kyle & Greipp, 1983). Haemostatic disorders are common and, in our study, abnormal bleeding was reported in the clinical histories of 28% of patients and was the presenting complaint in 3%. Cutaneous ecchymoses and purpura were the most frequently recorded, although clinically significant bleeding, particularly from the gastrointestinal and renal tracts, was common. Bleeding at sites other than the skin was significantly associated with prolongation of the PT, but was independent of the TT. This relationship was apparent in a previous study of 66 patients with AL-amyloidosis (Yood et al, 1983) and from a survey of case reports (Greipp et al, 1981). Prolongation of the PT in some individuals may therefore add to the haemostatic defect from vascular fragility, increasing the likelihood that bleeding may be clinically significant.

Importantly, this result shows the PT, but not the TT, to be a clinically useful predictor of bleeding tendency. At least one abnormality in the PT, aPTT or TT was found in 51% of our patients, although there was considerable variation in the nature and magnitude of the defects. This reflects, at least in part, the clinical heterogeneity of our patients and the frequent co-existence of AL-amyloidosis with other phenomena that may influence coagulation, such as liver disease, vitamin K malabsorption and disseminated intravascular coagulation (Kyle & Greipp, 1983). Inherited or coincidentally acquired factor deficiencies are also expected in a small number of individuals in such a large study group. Despite this, consistent patterns of abnormality in the clotting screen have emerged: prolongation of the TT occurred in 32% of patients and prolongation of the PT in 24%. These abnormalities co-existed in only 7% of patients, suggesting that they have distinct aetiologies. Prolongation of the aPTT was comparatively uncommon and, in most cases, was associated with prolongation of the PT. Similarly, thrombocytopenia was uncommon and could be explained in most cases by underlying myeloma. We have demonstrated that, in AL-amyloidosis, a prolonged TT was associated in most cases with prolongation of the RT and an increased fibrinogen Clauss assay, but was not associated with the presence of a paraprotein nor with a bleeding tendency. Furthermore, in the multiple linear regression, prolongation of the TT was independently associated with proteinuria, hypoalbuminaemia and the presence of hepatic amyloid infiltration. It has been reported previously that, in some lymphoproliferative disorders, a tumour product, unrelated to any paraprotein, may prolong the TT by direct inhibition of thrombin (Tefferi et al, 1990), and that this characteristically manifests with a normal RT and a significant bleeding tendency. The available laboratory and clinical data in this study indicate that this is unlikely in AL-amyloidosis. Instead, prolongation of the TT in our patients is more likely to be related to abnormal fibrin polymerization (Wisloff et al, 1984). Nephrotic syndrome was present in 28% of our patients and may itself cause prolongation of the TT and RT because of the synthesis of abnormally glycosylated fibrinogen and through an effect of the associated hypoalbuminaemia. In both cases, fibrin polymerization may be significantly impaired (Abshire et al, 1995; Toulon et al, 1995). The close associations in our patients between prolongation of the TT and, independently, both proteinuria and hypoalbuminaemia suggests that nephrotic syndrome is an important determinant of the TT, and it is likely that the associated dysfibrinogenaemia and hypoalbuminaemia are responsible. Finally, our finding of a further independent association between prolongation of the TT and amyloid liver disease suggests that, in addition, some of our patients may show defective fibrinogen synthesis as a result of hepatic infiltration. These conclusions differ from those of a previous study in which it was also noted that prolongation of the TT was usually accompanied by prolongation of the RT and that there was an association with nephrotic syndrome (Gastineau et al, 1991). It was concluded from mixing studies that a thrombin inhibitor was present,

q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 454±460

Haemostatic Disorders in AL-amyloidosis although the possibility of hypoalbuminaemia was not considered in these experiments. Evidence presented here favours a combination of dysfibrinogenaemia and hypoalbuminaemia as the cause of prolongation of the TT in most of our patients. In only 21% of our patients was a prolonged TT associated with a normal RT and, in these cases, we were unable to exclude a thrombin inhibitor. In contrast to the TT, prolongation of the PT in our patients, although predictive of bleeding, was not associated with any of the measures of renal function or with the total amyloid load or its anatomical distribution. The apparently idiosyncratic nature of PT prolongation and its association with bleeding therefore has important clinical implications, in that significant haemostatic abnormalities may arise in patients otherwise considered to have limited disease. Within our further study group, the majority of patients with a prolonged PT or aPTT had an associated reduction in FX:C. Of the 10 individuals with prolongation of the PT that could not be explained by subnormal FX:C, only two individuals had prolongation of the PT in excess of 3 s and, in both of these cases, mild factor VII deficiency was identified. Similarly, factor XII deficiency was identified in a single patient with unexplained prolongation of the aPTT. Because a full range of factor assays was not performed in all our patients, we were not able to assess the possible contribution made by abnormalities in factors other than X and V to the observed coagulation abnormalities. However, given the close association between subnormal FX:C and prolongation of the PT and aPTT, our data indicate that abnormality of FX:C is likely to be the principal determinant of these abnormal clotting times in AL-amyloidosis. Even if our observed reduction in FX:C was a manifestation of a wider defect in vitamin K metabolism, we would expect to observe a higher proportion of patients with prolongation of both PT and aPTT. Although widely perceived as being a rare defect (Glenner, 1977), the frequency of FX deficiency in our further study group was 14%. Previous reports of factor X deficiency in AL-amyloidosis are largely single case reports, although in one reported series of 95 patients a factor X level of less than 20 iu/ml was observed in six (7%) patients (Greipp et al, 1981), but not all patients in this group were tested. Currently accepted explanations suggest that FX is eliminated from the circulation by selective binding to amyloid deposits in the perivascular tissues of the spleen (Glenner, 1977), and this suggestion has been cited as the justification for attempts to correct the haemostatic defects in some amyloidosis patients with splenectomy (Greipp et al, 1981). Our results show that a simple adsorptive model for FX deficiency is almost certainly an oversimplification. In our patients with subnormal FX:C, there was a consistently higher FX:Ag, whereas simple adsorption of FX would be expected to result in concordant FX:C and FX:Ag. Furthermore, there was no difference in the frequency of splenic amyloid involvement between those patients with FX deficiency and those with normal FX, and splenic involvement did not correlate with coagulation abnormalities. It is significant that almost all the previous reports of FX deficiency in AL-amyloidosis have used functional assays

459

alone to assess FX levels. A comparison of FX:C and FX:Ag has been made in only one previous report (Fair & Edgington, 1985) and, in common with our findings, there was a wide variation in the comparative magnitudes of FX:C and FX:Ag. Our data indicate that most of our patients with apparent FX deficiency have a combined defect resulting from both reduction in FX:Ag and FX-specific activity. Although adsorptive elimination of FX from the circulation may occur, in most cases, it can only be a partial explanation for the functional deficiency. Discordance between the FX assays may indicate either a functional impairment of FX or the presence of an inhibitor. The consistent linearity observed in our FX:C assay at different plasma dilutions suggests that the latter is unlikely. Functional impairment of plasma FX may arise in AL-amyloidosis from a defect in post-translational modification, especially glycosylation (Sinha & Wolf, 1993), or from inactivation of the FX active site in the circulation by a soluble ligand such as a monoclonal antibody or antibody fragment. REFERENCES Abshire, T.C., Fink, L.K., Christian, J. & Hathaway, W.E. (1995) The prolonged thrombin time of nephrotic syndrome. Journal of Pediatric Hematology and Oncology, 17, 156±162. Fair, D.S. & Edgington, T.S. (1985) Heterogeneity of hereditary and acquired factor X deficiencies by combined immunochemical and functional analyses. British Journal of Haematology, 59, 235±248. Furie, B., Greene, E. & Furie, B.C. (1977) Syndrome of acquired factor X deficiency and systemic amyloidosis in vivo studies of the metabolic fate of factor X. New England Journal of Medicine, 297, 81±85. Furie, B., Voo, L., McAdam, K.P.W.J. & Furie, B.C. (1981) Mechanism of factor X deficiency in systemic amyloidosis. New England Journal of Medicine, 304, 827±830. Gastineau, D.A., Gertz, M.A., Daniels, T.M., Kyle, R.A. & Bowie, E.J.W. (1991) Inhibitor of the thrombin time in systemic amyloidosis: a common coagulation abnormality. Blood, 77, 2637. Gatel, A. (1998) AL amyloidosis combined with acquired factor V deficiency. Annals of Internal Medicine, 128, 604±605. Glenner, G.G. (1977) Factor X deficiency and systemic amyloidosis. New England Journal of Medicine, 297, 108±109. Greipp, P.R., Kyle, R.A. & Bowie, E.J.W. (1981) Factor X deficiency in amyloidosis: a critical review. American Journal of Hematology, 11, 443±450. Korsan-Bengtsen, K., Hjort, P.F. & Ygge, J. (1962) Acquired factor X deficiency in a patient with amyloidosis. Thrombosis et Diathesis Haemorrhagica, 7, 558±560. Kyle, R.A. & Greipp, P.R. (1983) Amyloidosis (AL): clinical and laboratory features of 229 cases. Mayo Clinic Proceedings, 58, 665±683. Liebman, H., Chinowsky, M., Valdin, J., Kenoyer, G. & Feinstein, D. (1983) Increased fibrinolysis and amyloidosis. Archives of Internal Medicine, 143, 678±682. McPherson, R.A., Onstad, J.W., Ugoretz, R.J. & Wolf, P.L. (1977) Coagulopathy in amyloidosis: combined deficiencies of factor X and IX. American Journal of Hematology, 3, 225±235. Perlin, E., Brakman, P., Berg, H.S., Kirchner, P.T., Moquin, R.B. & Astrup, T. (1971) Enhanced blood coagulation and fibrinolysis in a patient with primary amyloidosis. Thrombosis and Haemostasis, 26, 9±14. Rydh, A., Hietala, S.O., Ahlstom, K.R., Pepys, M.B. & Hawkins, P.N.

q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 454±460

460

A. D. Mumford et al

(1998) Serum amyloid P component scintigraphy in familial amyloid polyneuropathy: regression of visceral amyloid following liver transplantation. European Journal of Nuclear Medicine, 25, 709±713. Sinha, U. & Wolf, D.L. (1993) Carbohydrate residues modulate the activation of coagulation factor X. Journal of Biological Chemistry, 268, 3048±3051. Tefferi, A., Nichols, W.L. & Bowie, E.J.W. (1990) Circulating heparin-like anticoagulants: report of five consecutive cases and a review. American Journal of Medicine, 88, 184±188. Toulon, P., Frere, E., Bachmeyer, C., Candia, N., Blanche, P., Sereni,

D. & Sicard, D. (1995) Fibrin polymerisation defect in HIVinfected patients ± evidence for a critical role of albumin in the prolongation of thrombin and reptilase times. Thrombosis and Haemostasis, 73, 349±355. Wisloff, F., Michaelsen, T.E. & Godal, H.C. (1984) Inhibition or acceleration of fibrin polymerisation by monoclonal immunoglobulins and immunoglobulin fragments. Thrombosis Research, 35, 81±90. Yood, R.A., Skinner, M., Rubinow, A., Talarico, L. & Cohen, A.S. (1983) Bleeding manifestations in 100 patients with amyloidosis. Journal of the American Medical Association, 249, 1322±1324.

q 2000 Blackwell Science Ltd, British Journal of Haematology 110: 454±460