 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1997 The Journal of Bone and Joint Surgery, Inc.
Familial Heterozygous Protein-S Deficiency in a Patient Who Had Multifocal Osteonecrosis. A Case Report* , BALTIMORE , CINCINNATI, OHIO, BALTIMORE, MARYLANDInvestigation performed at Good Samaritan Hospital, Baltimore, and Jewish Hospital, Cincinnati
Osteonecrosis that is associated with the use of drugs such as corticosteroids14,24,32,35,39 or with underlying disease processes36,45,48,49 such as alcohol abuse39 has been characterized as secondary. Despite extensive investigations of the demographic characteristics of patients who have osteonecrosis32,35,39, as many as 39.5 per cent of these patients (934 of 2364 in one study39) have had no known risk factors, and the osteonecrosis has been characterized as idiopathic. However, as one of us (C. J. G.) and colleagues16-19,21-24,27 as well as others5,28,37,46 have shown, many instances of osteonecrosis that previously would have been regarded as idiopathic actually were associated with heritable thrombophilia (an increased tendency for intravascular thrombosis) and hypofibrinolysis (a reduced ability to lyse thrombi). The association between coagulation abnormalities and osteonecrosis has been recognized for more than thirty-five years2,28,37,41, but it has not been studied extensively, to our knowledge. Recent studies by one of us (C. J. G.) and colleagues16-24,27 and by others5,28,37,46 have redirected attention to this area. This association has major implications not only for the diagnosis and treatment of osteonecrosis but also for the screening of the relatives of affected individuals and the possible prevention of this and other thrombotic and thromboembolic disorders in those who are found to be at increased risk because of inherited thrombophilia or hypofibrinolysis17,18,21,22,24. We report the case of a patient in whom multifocal osteonecrosis was found to be associated with familial heterozygous protein-S deficiency, a thrombophilic disorder. To our knowledge, this is the first report in which a protein deficiency was found to be associated with osteonecrosis of the hip in an adult, although such deficiencies have been associated with Perthes disease18,22 and osteonecrosis of the jaw21,27. The purpose of the present report is to bring to the attention of the orthopaedic community the association between osteonecrosis and protein-S deficiency in this family.
A twenty-eight-year-old woman was seen because of a two-year history of pain in both shoulders, hips, and knees, as well as in the left ankle and foot, that had limited the activities of daily living. She was otherwise healthy and had no history of alcohol abuse, corticosteroid use, or other risk factors known to be associated with osteonecrosis. She was not taking birth control pills. The family history revealed that an uncle had died suddenly, presumably secondary to a pulmonary embolus. On musculoskeletal examination, she had a mildly antalgic gait on the left. In addition, she had bilateral pain in the groin as well as pain in the left foot and ankle with walking, mild discomfort in the shoulders with abduction and forward elevation, and moderate bilateral discomfort in the groin with any internal rotation of the hips. There was no notable limitation of motion except for that secondary to pain. Both knees had a full range of motion, complete ligamentous stability, and no effusion. Both ankles had a full range of motion, with pain of the left foot and ankle occurring only during walking. Plain radiographs of the shoulders revealed bilaterally symmetrical sclerotic lesions in the superomedial portions of both humeral heads without any subchondral radiolucency or irregularities of the joint surface. Magnetic resonance images demonstrated abnormalities in the same location; these findings were considered to be consistent with osteonecrosis. T2-weighted magnetic resonance images demonstrated a double-line sign35 (a band of increased signal intensity surrounded by a band of decreased signal intensity), which is considered pathognomonic of osteonecrosis. As there were no areas of collapse, these lesions were considered to be consistent with Ficat stage-II osteonecrosis14,15. Anteroposterior radiographs of the pelvis (Fig. 1-A) and frog-leg lateral radiographs of the hips showed a mottled pattern of increased radiodensity that emanated from the centers of the femoral heads and extended out to the articular surfaces, encompassing an arc of 120 degrees. Both hips had subchondral radiolucent areas with less than two millimeters of depression of the femoral head, findings that were consistent with Ficat stage-III involvement14,15. There was no involvement of the acetabulum or loss of the articular cartilage space. A 25-degree lateral margin of uninvolved bone, consistent with that noted in association with a medial or an Ohzono type-B lesion38, was observed in both hips. Magnetic resonance images showed lesions that encompassed approximately 50 per cent of the weight-bearing surfaces of both femoral heads (Fig. 1-B). Both hips had small effusions in the joint. Radiographs of both knees revealed normal findings. However, magnetic resonance images showed extensive involvement of both femoral condyles as well as both tibial plateaus (Fig. 2). Radiographs and magnetic resonance images of the left ankle and foot showed a small area of involvement in the proximal-lateral corner of the talus; the lesion appeared to be consistent with osteochondritis dissecans. The main body of the talus was not affected. However, the cuneiforms appeared to have extensive areas of osteonecrosis.
After the patient had fasted overnight, blood was obtained for the assessment of thrombophilia (as indicated by the levels of protein C, protein S, and anticardiolipin antibodies as well as by the resistance to activated protein C) and hypofibrinolysis (as indicated by the levels of lipoprotein[a] as well as by basal plasminogen activator-inhibitor activity and stimulated tissue-plasminogen activator activity) according to previously described techniques18,21,24,27. Measurements of fibrinolytic activity were normal: the level of lipoprotein(a) was ten milligrams per deciliter (normal, less than twenty-five milligrams per deciliter), basal plasminogen activator-inhibitor activity was 7.4 units per milliliter (normal, less than 26.9 units per milliliter), and stimulated tissue-plasminogen activator activity was 3.9 international units per milliliter (normal, more than 2.28 international units per milliliter). Most of the measurements of thrombophilia were normal: the resistance to activated protein C was 3.4 (normal, more than 2.0), the level of protein C was 92 per cent (normal, more than 70 per cent), the level of anticardiolipin IgG was sixteen units (normal, less than twenty-two units), and the level of anticardiolipin IgM was nine units (normal, less than ten units). The blood-chemistry profile, the hematocrit and the level of hemoglobin, the sickle-cell preparation, the erythrocyte sedimentation rate, and the lipid profile all were normal as well. When the initial sampling revealed that the patient had protein-S deficiency, and the results of other tests of coagulation were normal, the subsequent sampling of family members focused on protein S. None of the individuals tested were taking birth control pills, which have been associated with low levels of protein S16. The levels of total antigenic protein S and free protein S in the proband and in her first-degree relatives were compared with the values in the fifth to ninety-fifth percentile for sixty normal adults18, and the levels of functional protein S and C4b-binding protein were compared with the values in the fifth to ninety-fifth percentile for thirty normal adults18 (Fig. 3). With use of these cutoff points, the likelihood that the total, free, and functional protein-S levels of the proband were, by chance, below the fifth percentile for the normal controls was less than 5 per cent (p <0.05), a conventional marker of significance. Similarly, the fifth percentile for total protein S in thirty normal children18 was 76 per cent, and the likelihood that the level of total protein S of the proband's daughter was below this level by chance was less than 5 per cent. Because the proband and her family lived more than 7000 miles (11,265 kilometers) from our laboratory and all frozen samples of plasma had to be shipped by air, duplicate protein-S measurements were made only in the proband, in her daughter and son, and in a male sibling (Fig. 3).
Levels of total antigenic, free, and functional protein S and the level of C4b-binding protein were measured in two separate samples drawn from the proband, eight days apart (II-7, Fig. 3). Total antigenic protein-S levels also were measured in two separate samples from her protein S-deficient daughter and sister and from her normal son (III-1, II-4, and III-3, Fig. 3). In the proband, the total antigenic, free, and functional protein-S levels were below the fifth percentile for the normal controls (p <0.05). In the proband's daughter, the total antigenic protein-S level (60 and 55 per cent, as measured in two separate samples) was well below the fifth percentile (76 per cent18) for the normal pediatric controls (p < 0.05). The proband's daughter had a very low level of C4b-binding protein, which in turn may have exaggerated the level of free protein S6. The proband's affected sister, like the proband, had low total antigenic protein-S, free protein-S, and functional protein-S levels, all below the fifth percentile for the normal controls (p < 0.05 for all three levels). The proband's six normal first-degree relatives had predominantly normal total antigenic protein-S, free protein-S, and C4b-binding protein levels, within the fifth to ninety-fifth percentiles for the normal controls (Fig. 3). Alternative therapeutic modalities were reviewed with the patient. Core decompressions of both shoulders, hips, and knees were performed without complications. Histological analysis of all intraoperative specimens confirmed the diagnosis of osteonecrosis. Postoperatively, the patient participated in a program of physical therapy, walked with crutches and a four-point gait, and used a wheelchair for long distances; after six weeks, she progressed to full weight-bearing without restrictions. Anticoagulation therapy consisting of intravenous administration of heparin with conversion to oral administration of Coumadin (warfarin) also was initiated immediately postoperatively, and the dose was adjusted to maintain an international normalized ratio of 2.0 to 2.5.
There has been no consensus regarding the major pathophysiological mechanisms of osteonecrosis. Multiple theories have been advanced, with infarction35, fat embolism32,33, progressive ischemia30,31,47, and accumulative cell stress35 being suggested as mechanisms. Although the pathogenesis of idiopathic osteonecrosis remains to be fully explained, previous studies have demonstrated that coagulation disorders and thrombotic events may be pathogenic2,5,16-19,21-24,27,28,37,46. Jones32,33 reported that intravascular coagulation with fibrin-platelet thrombosis beginning in the vulnerable subchondral microcirculation (the capillary and sinusoidal bed) was especially associated with vasoconstriction and impaired secondary fibrinolysis (reperfusion of necrotic vessels with peripheral marrow hemorrhages) and that this appeared to be the final common pathway producing non-traumatic osteonecrosis. One of us (C. J. G.) and colleagues16-24,27 as well as others42,43, in published series of adults who had osteonecrosis of the hip or jaw and children who had Perthes disease, provided additional evidence to support this theory; in one study17, 220 (76 per cent) of 289 patients who had osteonecrosis were found to have thrombophilia or hypofibrinolysis, or both, primarily as autosomal dominant, highly penetrant heritable traits. Previous studies of children and adults who had osteonecrosis of the hip and adults who had osteonecrosis of the jaw have suggested that thrombophilia and hypofibrinolysis facilitate venous occlusion of the bone by fibrinous clots, leading to venous or sinusoidal hypertension within the cancellous bone17-19,21-24,27,34. Once the ischemic threshold is reached, cellular hypoxia presumably gives way to death of bone and marrow cells5,17-19,21-24,27. The detection of thrombophilia or hypofibrinolysis, or both, in 220 (76 per cent) of 289 patients who had osteonecrosis17 has substantial ramifications for the diagnosis and treatment of osteonecrosis as well as for the screening of family members and the prevention of thrombotic and thromboembolic events in those who are found to be at risk. These hypercoagulable states are predominantly inherited but also may be acquired16-19,21-24,27. The inherited disorders include deficiencies of specific protein inhibitors of the coagulation cascade (such as protein C, protein S, and structural abnormalities of factor V [resistance to activated protein C]3) and disorders of the fibrinolytic system (such as impaired release of tissue-plasminogen activator activity, increased levels of plasminogen activator-inhibitor, and high levels of lipoprotein[a]). Although predominantly inherited as autosomal dominant traits, these disorders also can occur de novo as a result of gene mutation26. There is also evidence that acquired coagulation disorders, when superimposed on underlying heritable disorders, can increase the likelihood of thrombosis synergistically21,23,27. One of us (C. J. G.) and colleagues demonstrated that exogenous estrogen supplementation (which by itself may induce resistance to activated protein C) synergistically increased the severity of osteonecrosis of the jaw in women who had underlying heritable resistance to activated protein C21,23,27. Familial protein-S deficiency was the only coagulation abnormality found in our patient, who had multifocal idiopathic osteonecrosis. She had no diseases that are associated with secondary osteonecrosis and was not taking medications that have been associated with that condition. Protein S is a vitamin K-dependent, antithrombotic plasma protein that serves as a cofactor for another antithrombotic plasma protein, protein C. Once activated, protein C inhibits the coagulation cascade by enzymatic cleavage of the activated forms of clotting factors V and VIII. Approximately 60 per cent of protein S is reversibly bound to C4b-binding protein and 40 per cent is free; when the concentration of C4b-binding protein increases the level of free protein S decreases, and when the concentration of C4b-binding protein decreases the level of free protein S increases6. Only the free form of protein S functions as a cofactor for activated protein C. Protein-S deficiency has been diagnosed on the basis of a reduction in the total antigenic protein-S level or a reduction in the free protein-S level with a low or below-normal total protein-S level6; both diagnostic criteria were met by the three affected family members in the present study (p < 0.05) (Fig. 3). The proband's daughter had low total protein-S levels (60 and 55 per cent) as well as a very low level of C4b-binding protein (58 per cent). The functional protein-S level was normal in the proband's daughter, probably because of the relatively high proportion of total protein S that was free, as the functional protein-S level reflects the free protein-S level and is based on enhancement of the anticoagulant effect of exogenous activated protein C by free protein S in the plasma8. One of us (C. J. G.) and colleagues recently reported that protein-S deficiency was detected in zero of thirty-one adults who had osteonecrosis17, in three (2.9 per cent) of 104 patients who had osteonecrosis of the jaw21,27, and in five (9.6 per cent) of fifty-two children who had Perthes disease18,22. The disorder was shown to be familial in two of the five kindreds in the latter two studies18,22. Deficiency of protein S has been associated with an increased tendency toward thrombosis1,4,7,9-13,29,40,42. This disorder is inherited as an autosomal dominant trait, with heterozygous individuals usually having plasma levels that are approximately one-half of the values in normal controls4,7. In the present report, two of the eight asymptomatic first-degree relatives of the proband (her six-year-old daughter and her thirty-two year-old sister) were found to have deficiencies of protein S and free protein S. In addition, the proband's uncle had died of a probable pulmonary embolus. In view of the inheritance pattern of these coagulation disorders (all of which are highly penetrant, autosomal dominant traits), we recommend the screening of all first-degree relatives of patients who have proved coagulation deficiencies. The prevalence of protein-S deficiency has been reported to be 8 per cent among patients who have thrombotic events before the age of forty years or who have a family history of thrombosis13,25. Our limited, non-blinded experience with eight patients16,17,19 suggests that it may be possible to retard or reverse the osteonecrotic process by correcting underlying coagulation disorders (such as thrombophilia and hypofibrinolysis) with use of stanozolol or Coumadin (warfarin). Such therapy needs to be initiated in the early stages of osteonecrosis (Ficat stage I or II14,15)16,17,19, before irreversible segmental collapse of the head of the femur occurs. Correction of thrombophilia or hypofibrinolysis may allow for the recovery of venous drainage and the reduction of intramedullary venous hypertension, thereby leading to normal oxygenation of bone and reversal of the ischemic necrosis of bone16,17,19. To our knowledge, this is the first report in which a protein-S deficiency was found to be associated with osteonecrosis of the hip in an adult. As this deficiency, as well as others in the coagulation pathway, recently was identified in 220 (76 per cent) of 289 patients who had osteonecrosis17, we believe that the systematic evaluation of coagulation and fibrinolysis is indicated for all patients who have osteonecrosis. Careful screening of the first-degree relatives of these patients also is indicated. The diagnosis of coagulation disorders may affect not only the diagnosis, treatment, and prevention of both idiopathic and secondary osteonecrosis but also the prevention of thrombotic and thromboembolic events in affected first-degree relatives. Advances in genetic mapping and engineering may allow for targeted treatment with manufacture of deficient proteins, with use of recombinant gene techniques. Because the associations between thrombophilic and hypofibrinolytic disorders and osteonecrosis of the jaw and the hip have been recognized only recently, there are as yet no data from placebo-controlled, blinded studies to indicate whether the treatment of such coagulation abnormalities has an effect on osteonecrosis. We have initiated such studies and encourage others to do so as well. However, the available data on eight adult patients16,17,19 suggest that the treatment of coagulation disorders does not confer a musculoskeletal benefit if it is initiated after the irreversible segmental collapse of the head of the femur (that is, when the patient has Ficat stage-III or IV involvement14,15) but that treatment can confer dramatic symptomatic and radiographic benefits if it is initiated well before the irreversible death of bone (that is, when the patient has Ficat stage-I or II involvement14,15). These findings emphasize the importance of the earliest possible diagnosis.
* No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Introduction
Case Report
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |