Dys®brinogenemia and Thrombosis Timothy Hayes, DVM, MD c Objectives.ÐTo review the state of the art relating to sus of experts attending the conference was reached. The congenital dys®brinogenemia as a potential risk factor for results of the discussion were used to revise the manuscript thrombosis, as re¯ected by the medical literature and the into its ®nal form. consensus opinion of recognized experts in the ®eld, and Conclusions.ÐConsensus was reached on 5 conclusions to make recommendations for the use of laboratory assays and 2 recommendations concerning the use of testing for for assessing this thrombotic risk in individual patients. dys®brinogens in the assessment of thrombotic risk in in- Data Sources.ÐReview of the medical literature, pri- dividual patients. Detailed discussion of the rationale for marily from the last 10 years. each of these recommendations is found in the text of this Data Extraction and Synthesis.ÐAfter an initial assess- article. Compared with the other, more common heredi- ment of the literature, key points were identi®ed. Experts tary thrombophilias, dys®brinogenemia encompasses a di- were assigned to do an in-depth review of the literature verse group of defects with varied clinical expressions. and to prepare a summary of their ®ndings and recom- Congenital dys®brinogenemia is a relatively rare cause of mendations. A draft manuscript was prepared and circu- thrombophilia. Therefore, routine testing for this disorder lated to every participant in the College of American Pa- is not recommended as part of the laboratory evaluation thologists Conference on Diagnostic Issues in Thrombo- of a thrombophilic patient. This is an evolving area of re- philia. Each of the key points and associated recommen- search, and further clinical studies may change these rec- dations were then presented for discussion at the ommendations in the future. conference. Recommendations were accepted if a consen- (Arch Pathol Lab Med. 2002;126:1387±1390) ibrinogen is a large (340 000 MW), complex molecule ther congenital or acquired in origin. Acquired dys®bri- F composed of 2 identical subunits linked through a di- nogenemias are associated with a heterogeneous group of sul®de bond. Each of the subunits consists of 3 polypep- underlying conditions, most commonly in the setting of tide chains (Aa,Bb,andg), which are encoded by 3 sep- liver disease. arate genes on the long arm of chromosome 4.1 The ®nal There are now in excess of 350 reported cases of unre- step in the coagulation cascade is the conversion of soluble lated families with hereditary (congenital) dys®brinoge- ®brinogen into insoluble ®brin polymer. This conversion nemia.2 Of these, the structural defects have been deter- is mediated by thrombin, which cleaves ®brinopeptide A mined in only 170.2±4 Most of these defects result from from the Aa chain and ®brinopeptide B from the Bb chain. DNA point mutations with single amino acid substitutions The release of ®brinopeptide A exposes new sites, which in the translated ®brinogen molecule. have a high af®nity for other ®brin molecules, leading to Given that ®brinogen plays a pivotal role in both the spontaneous ®brin polymerization. The release of ®bri- procoagulant and ®brinolytic pathways, it could be antic- nopeptide B is not required for ®brin polymerization. The ipated that defects in ®brinogen function might be asso- growing polymer is reinforced by the action of factor XIII ciated with increased risk for both hemorrhagic and (activated by thrombin), which covalently bonds the ends thrombotic disorders. Of the hereditary dys®brinogene- of ®brin strands together. The ®brin polymer is essential mias, several studies2,3,5 have shown that approximately for stabilization of the friable platelet plug, on which it is 55% are asymptomatic, 25% are associated with a bleeding deposited. Under physiologic conditions, the ®brin poly- tendency only, and the remaining 20% experience throm- mer is slowly digested by the ®brinolytic system. bosis with or without hemorrhage. As a group, the dys®brinogenemias are de®ned as dis- Before 1995, the relationship between venous thrombo- orders of ®brinogen structure, which may or may not re- sult in abnormal function. These abnormalities may be ei- sis and congenital dys®brinogenemia had been published only in case reports, small studies, and 9 relatively large retrospective studies. These latter reports included a total Accepted for publication June 17, 2002. of 2376 patients, with each study composed of 74 to 578 6±14 From the Spectrum Medical Group and Department of Pathology, patients. Collectively, these studies reported that pa- Maine Medical Center, Portland. tients with venous thrombosis had a 0.8% prevalence of Presented at the College of American Pathologists Consensus Con- congenital dys®brinogenemia.5 This information was in- ference XXXVI: Diagnostic Issues in Thrombophilia, Atlanta, Ga, No- suf®cient to con®rm a cause-effect relationship between vember 9±11, 2001 Reprints: Timothy Hayes, DVM, MD, Department of Pathology, dys®brinogenemia and thrombosis. This was addressed in Maine Medical Center, 22 Bramhall St, Portland, ME 04102 (e-mail: 1995 by a report from the International Society on Throm- [email protected]). bosis and Haemostasis Scienti®c and Standardization Arch Pathol Lab MedÐVol 126, November 2002 Dys®brinogenemia and ThrombosisÐHayes 1387 Table 1. Levels of Evidence for Consensus brinogenemias have also been associated with other clin- Recommendations ical conditions. The hereditary a®brinogenemias and dys- ®brinogenemias have historically been viewed as distinct Level 1: The recommendation is based on 1 or more well- designed prospective studies or 2 or more well-de- entities. It is becoming increasingly clear that the a®brin- signed retrospective studies ogenemias should be considered as a subset of dys®bri- Level 2: The recommendation is based on retrospective studies nogenemia in that several of these have now been shown or multiple anecdotal studies that reach consensus to result from structural defects.15 There is a high fre- Level 3: The recommendation is based on isolated anecdotal quency of spontaneous abortions, especially in patients studies and/or the consensus of expert practitioners with low levels of functional ®brinogen.16 Interestingly, 4 different ®brinogen mutations have been implicated in the pathogenesis of hereditary amyloidosis, with incorpora- Table 2. Conclusions tion of ®brinogen fragments into the amyloid ®brils.2 c The term dys®brinogenemia encompasses a variety of ab- PATHOPHYSIOLOGY normalities of ®brinogen function and may be either con- genital or acquired in origin.2,3,5 The exact mechanism by which dys®brinogenemia in- c Acquired forms of dys®brinogenemia appear to be weakly creases the risk for thrombosis is unknown and most like- associated, if at all, with thrombosis. ly varies, depending on the nature of the speci®c ®brino- c Using current laboratory methods, congenital dys®brinoge- gen defect. Theoretically, an abnormal ®brinogen might 5 nemia is found rarely in patients with venous thrombosis. lead to excessive thrombosis through either increased clot Level 2 c The relative risk of thrombosis is unknown, although there formation or impaired clot dissolution (®brinolysis). Liu et 17 is a signi®cant association between some forms of congen- al suggested that thrombosis might result from defective ital dys®brinogenemia and venous thrombosis. Level 2 binding of thrombin by the abnormal ®brinogen molecule. c The thrombin time is the most frequently used screening as- The excess circulating thrombin would therefore be free say to detect dys®brinogens in general, but it suffers from a to stimulate platelet activation. Defective thrombin bind- 2 signi®cant lack of sensitivity and speci®city. Level 2 ing has been demonstrated by several of the various dys- ®brinogens.18,19 Impaired ®brinolysis has also been re- ported, including resistance to lysis by plasmin20 and ab- Table 3. Speci®c Recommendations normal binding of tissue-type plasminogen activator.19 In only a few cases have the dys®brinogens been sig- c Although there is an association between some rare dys®- brinogens and thrombosis, testing for the identi®cation of ni®cantly correlated with increased risk for thrombosis. dys®brinogens in patients with thrombophilia is not recom- These include ®brinogens Marburg, Caracas V, Paris V, mended.5 Level 2 Chapel Hill III, Hannover II, Germany-k, New York I, Nig- c The diagnosis of inherited dys®brinogenemia associated megen, Ijmuiden, Christchurch II, Christchurch III, Lon- with thrombosis requires extensive analysis of the ®brinogen don VII, Vlissingen/Frankfurt IV, Melun, and Milano III.2 protein, the genetic defect, and family studies. Level 3 Although defects have been found throughout the ®brin- ogen molecule, these 15 ``thrombophilic'' dys®brinogens have mutations that are predominantly found in the C- Committee Subcommittee on Fibrinogen.5 The Working terminal domain of the Aa chain and the thrombin cleav- Party began by arbitrarily de®ning criteria for congenital age site of the Bb chain. Despite these recent advances in dys®brinogenemia, which included either demonstration our knowledge of the molecular basis of these dys®bri- of abnormal coagulation assays (ie, prolonged thrombin nogens, there is still no clear understanding as to how they time and a low functional-immunologic ratio of ®brino- may precipitate a thrombophilic state.5 It is now also clear gen) or detection of a ®brinogen defect using protein or that the clinical symptoms associated with the various DNA analysis. Twenty-six cases of dys®brinogenemia dys®brinogenemias may be modulated by coexisting fac- with thrombophilia were evaluated along with 187 rela- tors. For example, family studies suggest that being het- tives (99 had dys®brinogenemia and 88 did not). Twenty erozygous for both ®brinogen Cedar Rapids and factor V of the relatives with dys®brinogenemia experienced a Leiden results in thrombosis during pregnancy, whereas thrombotic event compared with none of those without family members with only one of these defects were dys®brinogenemia.