Considerations on a Tentative Classification of FVII Deficiency Suited for Practical Clinical Purposes

Abstract

Adequate classifications of disorders are of paramount importance in the management of congenital bleeding disorders. Classification of congenital FVII deficiency should be simple, based on few tests using thromboplastins of different origin. The first thromboplastin to be used is a rabbit brain preparation since it has been proven that this is the one that, overall, yields the lowest activity level. This is particularly so since molecular biology techniques have supplied important information with regard to the structure–function relation but have failed to supply a satisfactory classification of the defect. Mutations in the same domain have yielded different forms of FVII deficiency. Furthermore, molecular biology techniques are time consuming and are not feasible in every laboratory. A classification of FVII deficiency based on clinical, clotting, and immunological assays is proposed. This classification is suited for practical clinical purposes and may represent a useful preparatory basis for molecular biology studies.

Congenital bleeding disorders are a heterogeneous group of diseases which can be divided in X-linked, autosomal dominant, and autosomal recessive. The first 2, namely hemophilias and von Willebrand disease, represent about 90% of all bleeding disorders and are those most widely known. The third group, autosomal recessive, is a rare condition which, because of its rarity, is often considered less important. This is true if one takes into account but is not true from a scientific point of view. The discovery of prothrombin, factor VII, and factor X (FX) deficiencies has allowed the understanding of mechanisms involved in both the intrinsic or the extrinsic system of blood coagulation and has contributed a lot to our comprehension of the clotting event. The same is true for other bleeding conditions such as fibrinogen, FV or FXIII deficiencies. The main objective in the study of a congenital coagulation disorders is a comprehensive and suited classification for practical, clinical purposes.

Coagulation disorders may be simply classified on clinical grounds (severe or mild forms) or by means of clotting and immunological assays or by means of molecular biology techniques. The classical distinction between type I: cross-reacting material (CRM) negative or true deficiency and type II (CRM-positive forms) was made possible by the availability of suited precipitating antibodies against the single proteins. It was subsequently seen that the CRM-positive forms were also a heterogeneous group because sometimes these abnormal proteins behaved strangely in different clotting systems. The first example concerned a hemophilia B variant namely hemophilia BM. ¹ This peculiar disorder showed in fact a marked prolongation of the prothrombin time when the tissue factor used in the assay system was derived from ox brain.

The second one was FX Friuli, a CRM-positive variant characterized by discrepant results in the clotting assay. ² Abnormal in both the intrinsic and extrinsic assay system but normal when Russell’s viper venom was used in the assay system.

Both hemophilia BM variant, FX Friuli, and other abnormal forms received great acceptance. Today, it is customary to screen hemophilia B patients by carrying out, besides an activated partial thromboplastin time, an ox brain thromboplastic prothrombin time or a thrombotest. ³ The same is true for FX deficiency since at least 3 assays are usually used, namely an intrinsic system, an extrinsic system, and a Russell’s viper venom-dependent test.^4,5

This easy clotting diagnostic approach occurs often as a preliminary step before an immunological evaluation of the defect (enzyme-linked immunosorbent assay, immunoelectroforesis, etc) is carried out. It is clear in fact that these activating discrepancies can occur only if CRM is present.

The approach to FVII deficiency has not been similarly well established.^6,7 The existence of CRM-positive forms was first established in 1971 by Goodnight et al ⁸ and confirmed 1 year later by Denson et al. ⁹

Then, in 1978, the first disreactive FVII defect was discovered, namely FVII Padua. ¹⁰ Despite the fact that the demonstration of the existence of special forms of CRM-positive FVII deficiency dates back to about 3 decades ago, several articles are published even today in which the evaluation of the defect is carried out using a single thromboplastin, usually a rabbit brain thromboplastin, a human placenta reagent or, more recently, human recombinant preparations.^{11 –14}

This may be misleading since the FVII activity level found could not be the lowest or the highest observable in a given patient. For example, in FVII Padua (Arg304Gln) or in FVII Nagoya (Arg304Trp), the level obtained by rabbit brain is about 5% of normal, whereas that obtained using ox brain preparations is 100% of normal.^10,15

It is commonly stated that thromboplastins of human origin, for example, that obtained from human placenta or human recombinant thromboplastins should be used. This is based on the potential or real advantage obtainable in measuring human FVII. This may be true in some cases, for example in the control of coumarin treatment but is not absolute.

Again, FVII Padua and FVII Nagoya, using human thromboplastins, show levels of FVII activity of about 35% of normal. That level is about 4 or 5 times higher as that obtained with rabbit brain thromboplastins (about 5%), but about 3 times lower than that obtained using ox brain preparations.^10,15 In the study of congenital FVII deficiency, the first thromboplastin to be used is probably a rabbit brain thromboplastin preparation. In fact this thromboplastin is the one that yields always the lowest activity level even in CRM-positive variants.

The Arg304 residue seems a hot spot but is not the only area of FVII involved in binding of tissue factor.^{16 –19} At least, the Arg79 residue is also involved. In fact a few patients with the Arg79Gln mutation (FVII Shinjo) have also shown a pattern similar to that shown by FVII Padua (Arg304Gln), with nearly normal levels obtained using ox brain reagents. ¹⁸ Differences in reactivity toward thromboplastins of different origin have been described, at the heterozygous level, also for the Gly331Asp mutation; but in this case, discrepancies were less evident since one was dealing with heterozygotes. However, the highest level in this case is the one obtained using human tissue. ²⁰

For all the above-mentioned observations, it is clear that the type II FVII defects should be subdivided in at least 2 groups: those which contain an inert protein and those which contain a protein that is susceptible of partial or even complete activation given the presence of a peculiar tissue factor in the assay system. In this case, the protein on CRM could be termed as hypoactive or disreactive. Unfortunately, not all type II FVII defects have so far been investigated by means of a pool of thromboplastins of different origin. In reality, all these cases should be so investigated to find out whether their proteins are inactive or variably reactive. A possible candidate for the existence of thromboplastins-dependant discrepancies could be the Met298Iso mutation which shows low FVII activity using a rabbit brain thromboplastins but normal FVII antigen. ²¹

On the basis of the reports in the literature and on personal experience, the best and simplest approach is to use a panel of 3 tissue thromboplastins, namely a rabbit brain preparation, a human tissue (placenta) reagent, and an ox brain preparation.^{10,22 –24} Human recombinant reagents could also be used even if more expensive, in lieu of the human placenta reagent since it has been demonstrated that the values obtained using these 2 preparations are similar or even equivalent. ²⁵ The logical consequence of all the above-mentioned considerations justifies the following classification of FVII deficiency:

Classification of congenital FVII defects

Type I (true deficiency or CRM negative cases)

Type II (low activity, normal, near-normal or reduced antigen, or CRM-positive cases)

Forms with inert FVII CRM that shows no discrepancy in FVII activity (pseudo-true deficiency), regardless of the thromboplastin used

Forms with FVII CRM that shows different activity toward tissue thromboplastins of different origin.

Type III (cases of combined deficiency of FVII with other factors)

A comment has to be reserved for the combined defects of FVII. The combination, safe for other sporadic associations involves mainly FX.^{26 –28} There are 2 types of combined FVII and FX deficiency. The first type is due to a casual association of the 2 defects in the same family. ²⁷ In these cases, there is an independent segregation of the 2 defects in the same family. The second form is due to alteration (usually deletions) in chromosome 13 (q34) where both FVII and FX genes are located. ²⁸ In this condition, there is no independent segregation of the 2 defects in the family and the hereditary pattern is autosomal dominant and not recessive. Furthermore, these latter patients often also present variable malformations unrelated to blood coagulation such as mental retardation and bone abnormalities. ²⁶

Another association of FVII deficiency is that with bilirubin metabolism disorders. The association of FVII deficiency with Dubin Johnson syndrome seems casual because the 2 defects segregate independently. Furthermore, the gene for FVII and that for the multiple resistance protein 2 (MRP2) are located on different chromosomes, 13 and 10, respectively. The pattern is therefore different from that seen for combined FVII and FX (close genes on same chromosome 13). ²⁹

Once this basic clotting and immunological classification has been obtained, molecular biology techniques will allow to identify the site of the mutation (gla domain, epidermal growth factor [EGF] domain, catalytic domain, etc) and the type of mutations (nonsense, deletion, etc). However, the classification based on the domains of FVII is not needed for practical purposes since different phenotypes have been described for mutations in the same domain. For example, mutations in the catalytic domain may be associated with both type I or type 2 defects.^{6,10 –12,14}

Furthermore, it is surprising to realize that the same mutation (Arg to Gln) but in different areas of FVII, one in the catalytic domain of heavy chain (exon 8) and the other in the first EGF domain of the light chain, might yield the same pattern of activation.^{30 –33} The reason for this behavior is still being investigated.

It seems that the Arg79 residue of the EGF tethers mainly FVIIa to tissue factor, subsequently the Arg304 residue and, perhaps, other residues of the catalytic domain are required for complete catalytic activity of the bound FVIIa.^17,34

It is likely that, should the clotting and immunological approach underlined in this report be used, other peculiar variant cases could be discovered. This could also contribute to our understanding of the genotype–phenotype relation. Finally, and more importantly, this approach will have great bearing in deciding the substitution therapy. At least some of these CRM-positive variants, namely those which show a normal ox brain thromblastin test, are in fact asymptomatic or only paucisymptomatic and therefore need no or little substitution therapy at least in case of minor surgery.^10,24,35 This could eliminate the incidence of transfusion-related diseases and reduce costs. In particular, these patients should not be treated with Prothrombin Complex Concentrates (PCC) or activated FVII concentration for danger of thrombotic complications.^24,36 Several cases of thrombosis, mainly venous thrombosis, have been described in FVII-deficient patients ³⁷ and found to be due, in most cases, to the presence of associated risk factors . ³⁸

In conclusion, it is clear that an adequate classification of FVII defects suited for practical purposes may be obtained by means of simple tests which can be carried out in most clinical pathology laboratories.

Needless to say that molecular biology techniques should integrate the above-mentioned approach but they are not absolutely needed for an immediate and clinically satisfactory approach to the management of the patient.