Early Diagnosis of Ischemic Necrosis of the Femoral Head

David S. Hungerford

The Department of Orthopaedic Surgery, The Johns Hopkins University
School of Medicine and The Section of Orthopaedic Surgery, The Good Samaritan Hospital, Baltimore, Maryland

Hungerford, D. S. Early Diagnosis of Ischemic Necrosis of the Femoral Head. Johns Hopkins Medical Journal, 137 (1975), 270-275
©[Copyright Holder]. Reprinted with the permission of
The Johns Hopkins University Press.
All rights of reproduction in any form reserved.

Abstract: Intraosseous pressure in the intertrochanteric region, and response to a five milliliter saline load injected intraosseously were measured in twenty-four hips in twenty patients with ischemic necrosis of the femoral head and nine hips in seven controls. All hips in which subsequent biopsy proved ischemic necrosis of the femoral head showed either intraosseous pressure greater than 30 mmHg, a hypertensive response to the saline load, or both. All controls measure less than 30 mmHg intraosseous pressure and no significant pressure rise to the saline injection. Intramedullary venography in ischemic necrosis of the femoral head revealed poor filling of extraosseous veins, diaphyseal reflux and delayed clearance of dye. These changes were present in all stages of ischemic necrosis of the femoral head including the pre-radiologic stage and constitute the basis for early detection.

Just fifty years ago Haenisch (1) described the first case of idiopathic ischemic necrosis of the femoral head in an adult. By 1962, Mankin and Brower (2) were able to find only 22 cases in the English literature to which they added five. Since then, many publications, several with large series (3-5) have appeared, with Merle D'Aubigne (4) reporting a series of 104 patients with 150 hips showing ischemic necrosis of the femoral head (INFH).

Considerable controversy continues concerning etiology (6-9) and methods of treatment (3, 10, 11). However, there is general agreement regarding progression of the disease. Many patients, but by no means all, experience symptoms prior to the development of x-ray changes (4, 5, 12). Once x-ray changes have appeared, progression is the rule with collapse of the femoral head in the majority of patients within 6 months to 2 years following the onset of the disease (2, 4, 5). Although some might not consider patients with a history of rheumatoid arthritis or lupus erythematosis as suitable for including with "idiopathic" cases, the etiology of the femoral head osteonecrosis remains uncertain at best. In addition, review of such cases reveals a striking similarity in terms of symptoms, physical findings, x-ray changes, and progression of the disease compared to cases classified as idiopathic (13).

Serre and Simon (12) were the first to apply intraosseous phlebography to the proximal femur in INFH. After having demonstrated normal femoral arteriograms in six patients with INFH, they showed venous abnormalities consisting of venous stasis, diaphyseal reflux, and delayed intraosseous evacuation in all eleven cases so studied. One such case was in the preradiologic stage of the disease. At the time of venography, they also noted increased intraosseous pressure (IOP). Ficat and co-workers (3) have described increased intraosseous pressure and/or decreased outflow capacity in 100 cases of INFH, 42 in the pre-radiologic stage. Their classification is reviewed in Table I.

Although the incidence of bilateral involvement in reported series varies from 40 to 60% (2, 4, 5, 12, 14), Merle D'Aubigne (4) noted in his series that only 13% of patients presented initially with bilateral INFH while 50% eventually developed bilateral disease. This means that the majority of patients developing INFH on the second side will do so under the care of the physician treating the first side. Therefore a system of rapid, simple evaluation of the hemodynamic status of the suspicious hip is of considerable interest. Our present procedure and experience with such a system is the subject of this report.

X-Ray Classification of INFH
Joint Line
Osteoporosis Diffuse
Wedge Sclerosis
Early Collapse
Sequestrum Appearance
Marked Collapse
Extensive Destruction

The Suspicious Hip

Any patient with otherwise unexplained hip pain and with a history of one of the following must be considered as possibly having INFH and is suitable for study: (a) high alcohol intake; (b) high dose or prolonged steroid therapy; (c) rheumatoid arthritis or lupus erythematosis with or without steroid treatment; (d) significant trauma to the hip; (e) definite INFH on the contralateral side. Patients with primary diseases, requiring steroid therapy, seen by internists, rheumatologists, nephrologists, dermatologists, allergists, and pulmonary disease specialists will be subject to risk of developing INFH. The clinical presentation of INFH is quite variable but is best described by "root of the leg" pain-groin, proximal thigh or buttock in any combination. Occasionally, as with other hip disease, the symptoms are referred to the knee. The pain may be sudden or gradual in onset and initially may be evanescent. It is usually described as aching in nature. Physical examination inevitably reveals a painful limitation of motion with loss of internal rotation being most marked. Completely negative x-rays in no way rule out the diagnosis.

Method of Study

Under local or general anesthesia, sterile precautions, and biplane x-ray control or image intensification, a rigid, specially prepared needle is inserted percutaneously through the lateral femoral cortex at the level of the greater trochanter into the middle of the intertrochanteric area. The obturator is removed and with the aid of a 3-in., 22-gauge spinal needle, the intraosseous needle is filled with heparinized saline. The intraosseous needle is now connected via a three-way stop cock and saline-filled semirigid cannulae to a Validyne pressure transducer, and the pressure recorded on a Brush 250 pen recorder. Initial pressures often vary slightly and in most cases some minutes are required before a steady reading is obtained. The recording at 5 min is referred to as baseline pressure. At this point 5 ml of physiologic saline is injected intraosseously. The three-way stop cock is used in order to maintain a closed system for immediate post injection recording. The pressure recorded at 5 min after injection is referred to as the stress pressure (Fig 1.). In those cases where it is so desired, at the end of the recording 10 to 12 ml of soluble contrast media is injected through the recording needle and x-rays taken at the completion of injection and 5 min later. It should be noted that the intraosseous injection of contrast media under local anesthesia is uncomfortable in patients without INFH and excruciatingly painful in patients with INFH. Where intraosseous venography is carried out, systemic analgesia and sedation are necessary.


Twenty-four hips in 20 patients were hemodynamically evaluated. In each case, the patient had pain in the affected hip and biopsy confirmed the diagnosis of ischemic necrosis of bone. Patients with rheumatoid arthritis, lupus erythematosis, alcoholism, or steroid treatment were included. Patients with a history of trauma to the hip were excluded from this study. Ten patients had a history of high-dose steroid therapy while 10 patients had never had steroids. There was no difference in the findings, and they are presented as a single group. Table II shows the findings according to stage of disease. Nine hips in seven patients with normal x-rays, no symptoms or physical findings referable to the hip, or with normal biopsies and source of symptoms found elsewhere, served as controls. Intraosseous venography was carried out on five controls and seven patients with INFH. Venography in each of the control hips (Fig 2) was characterized by rapid filling of the three major sets of veins draining the hip, lack of diaphyseal reflux and complete or near complete clearing of dye on the 5-min film. Venographic changes in each of the seven cases with INFH so studied (Fig 3) were characterized in all stages by absent or incomplete filling of the main extraosseous veins, diaphyseal reflex, and stasis of contrast media as seen on the 5-min and subsequent films.

Fig. 1. Normal intraosseous pressure recording. (A) Opening pressure showing pulse pressure. (b) Baseline pressure at 5 min. (C) Recording five minutes after injection of five milliliters of saline intraaosseously. In this case the stress pressure (C) is actually marginally lower than the baseline pressure (B).

Intraosseous Pressure* Findings
# Hips
Baseline Pressure
Stress Test
17( 8-28)
18( 8-26)
*Figures in mmHg.

(A) Normal intraosseous venogram of proximal femur showing rapid and complete filling of A. circumflex Veins B. limino-capsular veins C. gluteal veins D. common femorall vein. (B) Film five minutes after injection demonstrates complete evacuation of radio-opaque material.

Eleven out of 24 pressure studies in hips with proven INFH showed baseline readings of 30 mm Hg or greater. Injecting 5 ml of saline intraosseously provoked an elevation of intraosseous pressure of more than 10 mm Hg in all 24 hips. This elevation persisted for longer than 5 min (Fig 4). These changes were irrespective of the stage of disease and in fact in this series more pronounced in Stage I (Table II). Note that there is no overlap in the range of stress test pressures between the subject and control groups. None of the controls showed baseline pressures greater than 30 mm Hg or positive stress tests.


Increased intraosseous pressure and venous stasis have been demonstrated in osteoarthritis by numerous authors (15-17). The degree of pressure elevation and venous stasis have been closely correlated with the degree of radiographic change (17). The findings in INFH are quite different, however, as several authors (3, 6, 7, 12) have noted, intraosseous hypertension and venous stasis in all stages of INFH, including a pre-radiologic stage. Although the precise role of these documented venous circulatory changes in the initiation and progress of INFH remain speculative, experimental production of venous obstruction has been demonstrated to produce increased osteoblastic and osteoclastic activity, as well as trabecular hypertrophy (18-20). In addition, Gourdou (13) was able to produce INFH in dogs by prolonged extraosseous elevation of venous pressure above 15 mm Hg in the lower extremity. With 2 hr of the production of increased extraosseous venous pressure unilaterally, the technicium 99 scan showed increased uptake on the affected side. Histology showed marrow necrosis and in the more severe cases death of osteocytes in trabeculae, findings similar to the microscopic changes in those cases of INFH diagnosed early.

(A) Intraosseous venogram in a patient with Stage I ischemic necrosis. Note the contrast between the normal appearing femoral head and the markedly deranged venogram showing no filling of the gluteal or limino-capsular systems, poor visualization of the circumflex veins, marked diaphyseal reflux with exit of contrast via diaphyseal veins. (B) Film 5 min after injection shows persistence of a large quantity of dye in the intramedullary space.

It is important to recognize that measurement of elevated IOP itself says nothing about blood flow in bone. In fact, its finding alone might be construed to indicate increased blood flow, as has been demonstrated experimentally by Azuma (21). On the other hand, Polster (22) has shown convincingly in dogs, that IOP is particularly sensitive to extraosseous venous pressure, with directly proportional changes. The association, therefore, of increased IOP with delayed clearance times of an injected load and finally with histologic evidence of marrow necrosis, marrow fibrosis, trabecular hypertrophy and death (Fig. 5) is convincing evidence for decreased tissue perfusion. The presence of these venous abnormalities can be simply documented by measurement of intraosseous pressure and intramedullary phlebography, and used in suspected cases to form the justification for conclusive establishment of diagnosis by biopsy.

Of course, the presence of increased IOP, positive stress test, and abnormal venography in Stages III and IV, does not contribute significantly to the diagnosis which is already established on the basis of characteristic radiologic changes. The presence of these findings, however, in all stages of the disease serves to underline the importance of these hemodynamic changes to the development and progression of the disease. Diagnosis in all cases must be confirmed by biopsy, but an index of suspicion in cases in Stage I or Stage II before the x-rays become characteristic can be elevated to a virtual certainly by the documentation of these intraosseous vascular changes.

Fig.-4. Intraosseous pressure recording in a patient with Stage I INFH. (A) Opening pressure; (B) baseline pressure; (C) stress pressure. Both baseline pressure and stress test are markedly abnormal. Note pulse pressure and in B and C reflection of respiration in the tracing.

Fig.-5. Photomicrograph from mid-portion of femoral neck in patient with Stage II INFH. Thickened trabeculae, lined with osteoblasts but with central areas devoid of osteocytes characterize the specimen. Marrow shows mild fibrosis.

Several ways of treating INFH by conservative surgical procedures have been reported. Merle D'Aubigne and co-workers (4) have reported on intertrochanteric osteotomy. Bonfiglio and Voke (10) and Marcus, Enneking, and Massam (11) have employed fibular-tibial strut grafts. Ficat and co-workers (3) have carried out core decompression. The common denominator of successful outcome from these surgical procedures has been early diagnosis. The patients in the high risk group, particularly those on steroids or with a history of alcoholism will be the primary beneficiaries. Given the high incidence of eventual bilaterality, these studies can be easily carried out on a contralateral hip at the time of an orthopaedic procedure on a unilateral advanced case of INFH. A high index of suspicion can lead to early diagnosis and the potential of salvage rather than eventual replacement.


  1. Haenisch: Arthritis dissecans der hufte. Zentralblatt fur Chir 52:999, 1925
  2. Mankin HJ and Brower TD: Bilateral idiopathic aseptic necrosis of the femur in adults: "Chandler's disease". Bull Hosp Joint Dis 23-24:42-57, 1962
  3. Ficat P, Arlet J, Vidal R, Ricci A, and fornial J: Resultats therapeutiques du forage biopsie dans les osteonecroses femoro-capitales primitive (100 cas). Rev Rhum 38:269-276, 1971
  4. Merle D'Aubigne R, Postel M, Mazabraud A, Massias P, and Gueguen J: Idiopathic necrosis of the femoral head in adults. J Bone Joint Surg 47-B612-633, 1965
  5. Patterson RJ, Bieckel WH and Dahlin David C: Idiopathic avascular necrosis of the head of the femur. J Bone Joint Surg 46-:267-282. 1964
  6. Arlet J, and Ficat P; Diagnostic de l'osteonecrose femoro-capitale primitive au Stade I (Stade preradiologic). Rev chir Orthop 54:637-648, 1968.
  7. Arlet J, Ficat P, and Sebbag D: Interet de la Mesure de la Pression intra-medullaire dans le massif Trochanterier chez l'homme en particulier pour le Diagnostic de l'osteonecrose Femoro-capitale. Rev Rhum 35:250-256, 1968
  8. Jones JP, Jr., and Sakovich Leo: Fat Embolism of Bone: A roentgenographic and histological investigation with use of intra-arterial lipidol in rabbits. J Bone Joint Surg 48-A:149-164, 1966
  9. Jung A, Kehr P, Hamid M, and Sall B: Pathogenie arterielle de l'osteonecrose primitive de la hanche. In La Circulation Osseuse. Arlet J and Ficat A. Eds. Paris: Inserm, 1973
  10. Bonfiglio M and Voke EM: Aseptic necrosis of the femoral head and non-union of the femoral neck. J Bone Joint Surg 50-A:48-66, 1968
  11. Marcus ND, Enneking WF, and Massam RA: The silent hip in idiiopathic aseptic necrosis. J Bone Joint Surg 55-A:1351-1366, 1973
  12. Serre H and Simon L: L'osteonecrose primitive de la tete femorale chez l'adulte. Acta Rheum Scan 7:265-286, 1961
  13. Gourdou J-F: Exploration Fonctionelle de L'Osteonecrose Post-Traumatique. Thesis, Toulouse, France, 1973
  14. Seze S, Welfing J, and Lequesne M: L'osteonecrose primitive de la tete femorale chez l'adulte. Revue du Rhum 27:117-127, 1960
  15. Arnoldi CC, Linderholm H, and Mussbichier H: Venous engorgement and intraosseous hypertension in osteoarthritis of the hip. J Bone Joint Surg 54-B:409-421, 1972
  16. Hulth, A: Circulatory disturbances in osteoarthritis of the hip. Acta Orthop Scan 28:81-89, 1959
  17. Lynch JA: Venous abnormalities and intraosseous hypertension associated with osteoarthritis of the knee. In The Knee Joint. Ingwersen OS, Van Linge B, Van Rens ThJG, Rosingh GE, Vraart BEEMJ, and La Vey D, Eds. Amsterdam: Excerpta Medica, 1974
  18. Bernstein M: Experimental production of arthritis by artificially produced passive congestion. J Bone Joint Surg 15:661-673, 1933
  19. Brookes M and Helal B: Primary osteoarthritis, venous engorgement, and osteogenesis. J Bone Joint Surg 50-B:493-504, 1968
  20. Rutishauser E, Rohner A, and Held D: Experimentelle Untersuchungen uber die Wirkungder Ischaemie auf den Knochen und das Mark. Virchows Arch Path Anat 333:101-118, 1960
  21. Azuma H: Intraosseous pressure as a measure of hemodynamic changes in bone marrow. Angiology 15:396-406, 1964
  22. Polster J: Zur Haemodynamik des Knochens. Stuttgart: Ferdinand Enke Verlag, 1970