TO THE EDITOR:
The etiology of Legg-Perthes disease is poorly understood, but recent reports have suggested a strong association with heritable thrombophilic defects. Glueck et al., in their study "Association of Antithrombotic Factor Deficiencies and Hypofibrinolysis with Legg-Perthes Disease" (78-A: 3—13, Jan. 1996), reported that 59 per cent (twenty-six) of forty-four children who had Legg-Perthes disease had a thrombophilic defect: nineteen (43 per cent) had protein-C deficiency, four (9 per cent) had protein-S deficiency, and three (7 per cent) had hypofibrinolysis. As part of an ongoing, unpublished study, we screened twelve white individuals who had Legg-Perthes disease for heritable thrombophilic defects. The ten male and two female patients had a median age of eleven years (range, six to forty-three years). None of the individuals had a personal or family history of thromboembolism.
We performed several investigations: coagulation screening (prothrombin time, activated partial thromboplastin time, and thrombin time), Clauss fibrinogen11, protein C (Coamatic; Chromogenix, Mölndal, Sweden), free and total protein S (enzyme-linked immunosorbent assay; Dako, Glostrup, Denmark), antithrombin (Coamatic; Chromogenix), resistance to activated protein C by a modified activated protein-C resistance assay with factor-V deficient plasma (Coatest; Chromogenix), plasminogen (Coamatic; Chromogenix), fibrin plate lysis, tissue-plasminogen activator antigen (Immulyse; Biopool, Umea, Sweden), plasminogen activator inhibitor (Coatest; Chromogenix), and factor XII (Unichrom; Rho Diagnostics, Gerrards Cross, United Kingdom). Confirmatory DNA testing was performed on patients who had resistance to activated protein C on coagulation testing by polymerase chain-reaction screening for the factor-V Leiden mutation2. Levels of thrombin-antithrombin III complexes, a marker of thrombin generation, were also measured (Enygnost; Behring, Marburg, Germany).
One of the twelve patients had a low resistance to activated protein C with a ratio of 1.06 (normal, 2.1 to 4.0) associated with homozygosity for the factor-V Leiden mutation. Another had an elevated thrombin-antithrombin-III complex level, which, although not a specific thrombophilic defect, is indicative of increased thrombin generation. The remainder of the results for the thrombophilia screening were normal. We have not found as high a prevalence of thrombophilic abnormalities as that reported by Glueck et al. Our findings suggest a prevalence of detectable hemostatic defect of two of twelve, with a specific association with resistance to activated protein C and factor-V Leiden mutation. The allele frequency of factor-V Leiden in the normal population of the United Kingdom is 1.75 per cent1 compared with one of twelve individuals in our study. Glueck et al. did not include resistance to activated protein C in their thrombophilia screen. Because of the potential for antithrombotic therapy to limit sequelae of Legg-Perthes disease, it is important to establish the prevalence of thrombophilic defects in patients who have this disorder in the United Kingdom. This can be achieved only by the performance of a larger study, which will establish the value of routine thrombophilia screening in this group of patients.
D. M. Hunt, F.R.C.S.; Z. Holmes, Ph.D.; W. Pickering, F.I.B.M.S.; H. Cohen, M.D., F.R.C.P., F.R.C.Path.: Departments of Orthopaedics (D. M. H.) and Haematology (Z. H., W. P., and H. C.), Imperial College School of Medicine at St. Mary's Hospital, Praed Street, London W2 1NY, United Kingdom
Dr. Glueck, Dr. Crawford, Dr. Roy, Dr. Freiberg, and Mr. Stroop reply:
Thrombophilia may cause thrombotic venous occlusion in the femoral head, leading to venous hypertension and hypoxic death of bone in children (Legg-Perthes disease)7-9. In our first two evaluations of eight7 and thirty-three children (the present report) who had Legg-Perthes disease, we did not measure resistance to activated protein C or screen for the mutant factor-V Leiden gene because the assays were not then available to us. After we subsequently set up these assays and evaluated sixty-four children who had Legg-Perthes disease8,9, we reported that the mutant factor-V Leiden gene was the most common heritable thrombophilic trait in children and adults who had osteonecrosis, a finding that is consistent with its being the most common heritable thrombophilic trait in adults10 and children8. Genomic deoxyribonucleic acid was studied8 to delineate the CGA?CAA substitution at position 1691 of the factor-V Leiden gene responsible for resistance to activated protein C. Eight (13 per cent) of the sixty-four children who had Legg-Perthes disease had both a low activated protein-C ratio and the mutant factor-V gene (seven heterozygotes and one homozygote) compared with seven (4 per cent) of 169 healthy normal pediatric controls (chi square = 5.38, p = 0.02). Two or three-generation vertical and horizontal transmission of heterozygosity for the mutant factor-V gene was found in four of the eight kindreds8. In the two largest three-generation kindreds studied, seven of sixteen and twelve of sixteen family members evaluated were heterozygous for the mutant factor-V Leiden gene. The child probands who had Legg-Perthes disease in these two kindreds were homozygous and heterozygous, respectively, for the mutant factor-V Leiden gene8. In one of the eight kindreds, Legg-Perthes disease developed nine months later in the proband's brother, who was found to be heterozygous for the factor-V Leiden gene mutation and who was asymptomatic at the time of the initial sampling. Legg-Perthes disease later developed in a third sibling at the age of two years; this sibling was found to be heterozygous for the mutant factor-V Leiden gene. Heterozygosity or homozygosity for the mutant factor-V Leiden gene with thrombophilic resistance to activated protein C appears to be an important, pathogenetic cause of Legg-Perthes disease, and this information expands the understanding of the thrombophilic and hypofibrinolytic pathoetiology of this disease7-9.
The major promise arising from the documentation of coagulation abnormalities in the pathogenesis of osteonecrosis should lie in the fact that medical correction of these abnormalities may stop progression or even allow regression of osteonecrosis, not only in adults, as we have already shown4,5,9, but also possibly in children. In studies of osteonecrosis of the hip in adults, provided that therapy was begun before irreversible segmental collapse of the head of the femur4,5,9, treatment of thrombophilia and hypofibrinolysis was able to retard or reverse the progression of osteonecrosis. Two adults who had early, potentially reversible osteonecrosis of the hip (Ficat stages3 I and II) had thrombophilia (resistance to activated protein C and protein-C deficiency) and were managed with Coumadin (warfarin) (targeted international norm reference of 2 to 2.5) or six milligrams of Winstrol (stanozolol, an anabolic-androgenic steroid) a day4,9. Thrombophilia was normalized, severe pain in the hip resolved, and the progression of the osteonecrosis was reversed or retarded, as demonstrated on magnetic resonance images and radiographs. Two adults who had osteonecrosis of the hip (Ficat stages I and II) had hypofibrinolysis (high plasminogen activator-inhibitor activity and high levels of lipoprotein[a]). Six milligrams of Winstrol a day normalized the fibrinolytic activity, resolved the severe pain in the hip, and reversed or retarded the osteonecrosis, as demonstrated on magnetic resonance images and radiographs4,5,9. However, once segmental collapse of the femoral head with osteonecrosis has occurred (Ficat stages III and IV), it appears that correction of the underlying pathoetiological coagulation disorder will not stop the progression of the osteonecrosis4,5,9. Hence, in four hypofibrinolytic adults (high plasminogen activator-inhibitor activity and high levels of lipoprotein[a]) who had more advanced osteonecrosis (segmental collapse of the femoral head; Ficat stages III or IV), six milligrams of Winstrol a day normalized the fibrinolysis but had no effect on the pain in the hip or the findings on the magnetic resonance images or radiographs4,5,9. We postulate that treatment4,5,9 of thrombophilia and hypofibrinolysis early in the course of Legg-Perthes disease might allow effective treatment of the disease, thereby avoiding the all-too-common necessity for hip replacement in young adulthood. Given the promise of medical intervention in Legg-Perthes disease4,5,9, it will be important to characterize and stage the degree of involvement of the femoral head with use of magnetic resonance imaging and radiographs and to collect relevant clinical data (such as age, race, gender, age at the onset of pain, exposure to second-hand cigarette smoke6, family history of osteonecrosis, and so on) uniformly to systematize the reporting of the efficacy or absence of efficacy of treatment.
NOTE: The authors thank Ralph Gruppo, M.D., for assistance in preparing this reply.
C. J. Glueck, M.D.; Richard Freiberg, M.D.: Cholesterol Center (C. J. G.) and Department of Orthopedics (R. F.), Jewish Hospital, 3200 Burnet Avenue, Cincinnati, Ohio 45229
Alvin Crawford, M.D.; Dennis Roy, M.D.: Department of Orthopedics, Children's Hospital, 240 Bethesda Avenue, Cincinnati, Ohio 45229
Helen Glueck, M.D.: Deceased
Davis Stroop, M.S.: Departments of Pathology and Laboratory of Medicine, University of Cincinnati College of Medicine, University Hospital, 234 Goodman Avenue, Cincinnati, Ohio 45267