124

CLINICAL ORTHOPAEDICS AND RELATED RESEARCHNumber 385, pp. 124–129© 2001 Lippincott Williams & Wilkins, Inc.

This is the seventh reported case of recurrentLegg-Calvé-Perthes disease. The report docu-ments the initial onset in a boy 4 years of agewith healing clinically and radiographically.The boy experienced recurrence of disease at 8years of age with last followup at 20 years of age.Tests related to blood hypercoagulability andhypofibrinolysis and certain genetic factors rel-evant to osteonecrosis of the bone are reported.

A seventh case of recurrent Legg-Calvé-Perthesdisease is reported. In addition to a followup toskeletal maturity, an analysis was done ofthrombophilic, hypofibrinolytic, and moleculargenetic factors.

Ischemia is regarded as essential in the patho-genesis of this disorder.5 Two basic hypothesesattempt to explain the mechanism of the isch-emia. Either tamponade of the retinacular ves-sels occurs by increased intracapsular pres-sure,26 or cessation of blood flow to the capitalepiphysis occurs from an intravascular clot.Several studies support the latter theory.7,15,21

The authors present a case report of recur-rent Legg-Calvé-Perthes disease with long-termfollowup and additional evidence supportive ofa genetic predisposition for the disorder.

CASE REPORT

A 4-year 2-month old boy initially was seen at theauthors’ institution on March 11, 1981, for pain inthe left knee with a limp of 1-week duration. Onphysical examination, the range of motion (ROM)of the right and left hips was 140� flexion, 0� ex-tension, 60� abduction, 50� (right) and 40� (left) in-ternal rotation, and 85� (right) and 75� (left) exter-nal rotation. Findings seen on radiographs wereconsistent with avascular necrosis of the left capi-tal epiphysis with collapse and sclerosis of the an-terior 1⁄2 of the femoral head without epiphysealplate or metaphyseal involvement. The right hipappeared normal (Fig 1). There was no clinical, ge-netic, or radiologic evidence of skeletal dysplasias.The patient was treated with an abduction brace fit-ted in 45� abduction with instructions to wear thebrace night and day. Radiographs and clinical ex-aminations were done every 4 months (Fig 2).Range of motion and containment were adequate.Complete reconstitution of the superolateral frag-ment was seen 10 months later, and the child wasweaned from wearing the brace. A final clinicalcheck 17 months after onset of the disease showedcomplete healing of the capital epiphysis with fullmotion at the hips and no limp (Fig 3).

The patient presented again to the authors’ insti-tution 3 years later with left knee pain and a limp of2 weeks’ duration. Physical examination of the left

Recurrent Legg-Calvé-Perthes Disease Case Report and Long-Term Followup

David B. Stevens, MD*; Stanley S. Tao, MD*; and Charles J. Glueck, MD**

From the *Shriners Hospital for Children, Lexington,KY; and **The Cholesterol Center at The Jewish Hospi-tal of Cincinnati, Cincinnati, OH.Reprint requests to David B. Stevens, MD, Shriners Hos-pitals for Children, 1900 Richmond Road, Lexington,KY 40502–1298.Received: June 29, 1999. Revised: March 7, 2000; June 5, 2000; July 17, 2000.Accepted: August 1, 2000.

hip revealed 10� flexion, 60� abduction, 0� internalrotation, and 30� external rotation compared withnormal ROM of the right hip. Radiographs revealed4 mm of increased joint space on the left comparedwith the right with a subchondral fracture involving50% of the capital epiphysis (Fig 4). The patientagain was treated with an ambulatory, abductionbrace. Followup radiographs at 4-month intervalsshowed adequate containment and motion. This oc-currence was manifested by more extensive in-volvement of the femoral head with additional

Number 385April, 2001 Recurrent Legg-Calvé-Perthes Disease 125

Fig 1A–B. (A) Lateral radiograph on presenta-tion shows collapse of the anterior 1⁄2 of the leftfemoral head. (B) Anteroposterior radiograph re-veals some irregularity of the slope of the left cap-ital femoral epiphysis.

Fig 3A–D. (A) Anteroposterior view of normalright hip. (B) Anteroposterior view of left hipshows reconstitution of normal contour and tra-becular pattern of bone. (C) Right hip in lateralview remains normal. (D) Left hip in lateral viewshows the femoral head to be almost completelyrestored.

Fig 2. Six months after onset of disease, the an-terior 1⁄2 of the left capital femoral epiphysis re-mains collapsed and sclerotic.

A

B

A

C

B

D

pathologic changes in the physis and the metaphysis(Fig 5). The opposite hip remained normal.

At the most recent followup 12 years after the re-currence, the patient had excellent ROM of bothhips. Both hips had 135� flexion, 60� abduction, and75� external rotation. The right hip had 50� internalrotation and the left hip had 40� internal rotation.

The left lower extremity was 2.5 cm shorter than theright lower extremity measured from the xiphoidprocess to the medial malleolus of each ankle. Ra-diographs showed a coxa magna of the femoral headwith shortening of the femoral neck (Fig 6).

The patient agreed to have blood tested for ab-normal coagulation factors. The results are shownin Table 1. Results of deoxyribonucleic acid(DNA) analysis of genetic factors are shown inTable 2.

DISCUSSION

Legg,20 Calvé,5 Perthes,24 and Waldenstrom27

independently described this hip disorder inyoung children. Since then, only six reportedcases of recurrent Legg-Calvé-Perthes disease

Clinical Orthopaedics126 Stevens et al and Related Research

Fig 6A–B. (A) Anteroposterior view of pelvis 16years after initial onset and 12 years after recur-rence shows the round left femoral head al-though enlarged and with shortened left femoralneck. (B) Lateral view at the same time confirmsthe shortened left femoral neck with secondaryprominence of the greater trochanter.

Fig 5. Two years later, continued collapse withsclerosis of the femoral head is seen concurrentwith metaphyseal and physeal involvement.

Fig 4A–B. (A) Right hip in lateral view remainsnormal at age 8 years 7 months. (B) Lateral viewof left femoral head shows a clear crescent signfrequently associated with early onset of capitalfemoral head necrosis.

A B

A

B

after complete healing of the initial insult havebeen reported.3,4,15,16,19

The differential diagnosis is important inreporting recurrent disease after compete heal-ing. Legg-Calvé-Perthes disease remains aclinical diagnosis with variable symptoms andphysical findings and a wide spectrum of radi-ologic findings. Patients with Perthes diseasewho are younger than 4 years tend to havefewer symptoms and findings with lesser in-volvement as measured by the severity ofchanges on the radiograph. Systemic condi-tions are best ruled out by the genetic history,physical examination, and a skeletal radio-logic survey. In the current patient, these find-ings all were negative. Meyer’s dysplasia alsowas considered. Originally described by Ped-

ersen23 and Meyer22 as a subset of Perthes, itis manifested by earlier onset, fragmentationof the epiphysis, and a benign course. Diagno-sis of this disease was discovered accidentallyin 70% of the patients because they had nosymptoms. Meyer states that both conditionsare circulatory disturbances, with presentationdetermined by the age of the patient and per-haps by the severity of the vascular insult. Thereports of Khermosh and Wientroub18 andHarel et al11 do not add anything to the origi-nal work of Pedersen23 and Meyer.22 Contraryto these four reports,11,18,22,23 the capital fem-oral epiphyses in the patient reported in thecurrent study were of normal size and radio-logic density. The patient did not have theclinical findings of Meyer’s dysplasia.

Martinez and Weinstein21 reported the lastdocumented case of recurrent Legg-Calvé-Perthes disease in 1991. Their case was pre-sented before the outcome was known. Mar-tinez and Weinstein reported that all sixprevious cases had age of onset of the initialinsult between 2 and 6 years (average, 4years). The current patient initially presentedat 4 years of age. The previous patients alsohad an age range of 2.5 to 6 years (average, 4years) between the first episode of necrosisand recurrence after complete reossification.The current patient had repeat necrosis of thecapital epiphysis 3 years after complete heal-

Number 385April, 2001 Recurrent Legg-Calvé-Perthes Disease 127

TABLE 1. Blood Elements Related to Clotting

Test Normal Values PatieValues

Resistance to activated protein-C � 1.8 2.93Protein C � 70 111Protein S � 70 122Free protein S � 65 Not doneAntithrombin III � 80 Not doneHomocysteine 13.1 �mol/L Not doneMethylmalonic acid � 250 Not donePlasminogen activator inhibitor (PAI-fx) � 26.9 �/mL 15Stimulated tissue plasminogen activator activity � 2.28 IU/mL 2.7Serum insulin � 20 �U/mL 82.5Lipoprotein(a) � 25 mg/dL 1Triglycerides Not doneTotal cholesterol with high- and low-density lipoproteins Not done

TABLE 2. Deoxyribonucleic AcidAnalysis

Factor Factor in Patient

Plasminogen Activator 4G/4G homozygousInhibitor-1 Promotor (4G/5G) Polymorphism

MTHFRS (C677T) gene genotype normalmutation

Factor V (Leiden) mutation genotype normalProthrombin 3� untranslated genotype normal

region mutation

ing. Complete healing of the initial necrosiswas documented in all patients. The side af-fected in all patients described, including thecurrent patient, was the left side in three pa-tients and the right side in four patients. Sixmale patients and one female patient havebeen reported. Four of the seven patients hadbilateral disease, with the recurrent side hav-ing much less necrosis on initial presentation.Pertheslike changes occur in many syndromesand metabolic disorders, but no other condi-tions were present in the current patient or inhis family.

The second episode in previous cases wasreported to be more severe with more involve-ment of the femoral head compared with theinitial episode. This also was the case with thecurrent patient. Martinez and Weinstein21

speculated that this, in addition to the patientsbeing older at the time of the second insult,would lead to a worse prognosis. Long-termstudies of single-episode Legg-Calvé-Perthesdisease have found good function into thethird and fourth decades of life, even with de-formity of the femoral head. The current pa-tient, now 9 years after complete healing of thesecond insult, has no limitations in functionand is pain free. However, radiographs show aflattened femoral head with an associatedshortened femoral neck (Fig 6).

The exact etiology of Legg-Calvé-Perthesdisease is unknown. Some studies10,26 supporttheories involving vascular insufficiency. Anextracapsular arterial ring, ascending cervicalretinacular branches, and the artery of the liga-mentum teres provide blood to the femoral cap-ital epiphysis in the child. Some authors postu-late the disease is secondary to multiple infarctsof varying age. Sanchis et al25 first describedthe second infarction theory in 1973. Inoue etal14 confirmed the theory with histologic analy-sis. The multiple infarct theory lends itself wellto recurrent Legg-Calvé-Perthes disease.

An increased tendency to form blood clots(thrombophilia), a decreased ability to dis-solve blood clots (hypofibrinolysis), or bothcould interfere with the blood supply to thefemoral capital epiphysis. Boettcher et al2

identified sludging and thrombosis in someadult patients with nontraumatic necrosis ofthe femoral head. Kleinman and Bleck19 foundstatistically higher blood viscosities in pa-tients with Legg-Calvé-Perthes disease com-pared with healthy subjects. Gragosiewicz etal10 documented an increased level of alphaone-antitrypsin in their patients with Legg-Calvé-Perthes disease compared with controlsubjects. This could lead to a decrease in fibri-nolytic activity and a predisposition to in-creased intravascular clotting.

Several authors have reported alteration infactors that contribute to thrombophilia andhypofibrinolysis in patients with Perthes dis-ease.7,8,10,12 However, other investigators havenot been able to duplicate these findings intheir series.1,6

The current patient’s blood was tested forthrombophilic factors and hypofibrinolyticfactors. Thrombophilic factors that weretested included resistance to activated protein-c, protein-c, protein-s, and antiphospholipidantibodies. Results for these factors were all inthe normal range (Table 1). Hypfibrinolyticfactors tested included plasminogen activatorinhibitor (PAI-fx), stimulated tissue plasmino-gen activator activity (sTPA-Fx), and lipopro-tein (a). High PAI-fx often is associated withhigh levels of serum insulin. The current pa-tient’s serum insulin was 82.5 �U/mL (uppernormal 20 �U/mL). A high level of inhibitorto plasminogen activator theoretically shouldlead to decreased ability to lyse blood clots.This is in agreement with other reports, whichcite clotting factors and increased blood vis-cosity as potential causes of Legg-Calvé-Perthes disease.2,17 Usually, higher PAI-fx iscaused by a hereditary defect in the PAI gene.The patient’s DNA was analyzed for the genesassociated with thrombophilia and hypofibri-nolysis. This was done with DNA polymerasechain reactions to evaluate polymorphisms ofthe genes. The patient is homozygous for the4G/4G polymorphism in the promoter regionof the PAI-11 and has high serum insulin lev-els. The MTHFR (C677T) gene mutation, theFactor V (Leiden) mutation, and the pro-

Clinical Orthopaedics128 Stevens et al and Related Research

Number 385April, 2001 Recurrent Legg-Calvé-Perthes Disease 129

thrombin 3� untranslated region mutation werenormal.

This is the seventh documented case of re-current Legg-Calvé-Perthes disease. Followup9 years after healing of recurrence of diseaseshows the patient to be pain free with no func-tional limitations. Recent reports by Glueck etal,8,9 Hunt et al,13 Herndon,12 and Arruda et al1regarding the thrombophilic and hypofibri-nolytic factors and their contribution to Legg-Calvé-Perthes disease are conflicting, and thesignificance of these factors has not been de-termined, although the hypothesis that a ge-netic defect in the blood, which could lead tosludging of blood and osteonecrosis, is inter-esting. The results in the current patient arepresented for documentation and speculationconcerning their significance.

References1. Arruda VR, Belangero WD, Ozelo M, et al: Inherited

risk factors for thrombophilia among children withLegg-Calve-Perthes disease. J Pediatr Orthop19:84–87, 1999.

2. Boettcher WG, Bonfigilo M, Hamilton HH, SheetsRF, Smith K: Non-traumatic necrosis of the femoralhead: Relation of altered hemostasis to etiology. JBone Joint Surg 52A:312–321, 1970.

3. Burkhead WZ, Wenger DR: Recurrent Legg-Calve-Perthes disease. Iowa Orthop J 1:46–49, 1981.

4. Caffey J: Pediatric X-Ray Diagnosis. A Textbook forStudents and Practitioners of Pediatrics, Surgery andRadiology. Ed 4. Chicago, Year Book Medical968–973, 1961.

5. Calve J: Sur une forme particuliere de pseudo-coxal-gie greffee sur des deformations caracteristiques de l’extremite superieure du femur. Rev Chir 42:54–84,1910.

6. Catterall A: Legg-Calve-Perthes syndrome. Clin Or-thop 158:41–52, 1981.

7. Gallistl S, Rectinger T, Linhart W, Muntean W: Therole of inherited thrombotic disorders in the etiologyof Legg-Calve-Perthes disease. J Pediatr Orthop19:82–83, 1999.

8. Glueck CJ, Crawford A, Roy D, et al: Association ofantithrombotic factor deficiencies and hypofibrinol-

ysis with Legg-Perthes Disease. J Bone Joint Surg78A:3–13, 1996.

9. Glueck CJ, Glueck H, Mieczkowski L, et al: Famil-ial high plasminogen activator inhibitor with hy-pofibrinolysis: A new pathophysiological cause ofosteonecrosis? Thromb Haemost 69:460–465, 1993.

10. Gragosiewicz A, Okonski M, Stolecka D, et al: Isch-emia of the femoral head in Perthes’ disease: Is thecause intra- or extravascular? J Pediatr Orthop9:160–162, 1989.

11. Harel L, Kornreich L, Ashkenazi S, et al: Meyer dys-plasia in the differential diagnosis of hip disease inyoung children. Arch Pediatr Adolesc Med153:942–945, 1999.

12. Herndon WA: Letter to the Editor. J Bone Joint Surg79A:1114–1115, 1997.

13. Hunt DM, Holmes Z, Pickering W, Cohen H: Letterto the Editor. J Bone Joint Surg 80A:604–606, 1998.

14. Inoue A, Freeman MAR, Vemon-Roberts B, et al:The pathogenesis of Perthes disease. J Bone JointSurg 58B:453–461, 1976.

15. Kahle WK: Letter to the Editor. J Bone Joint Surg79A:1114–1115, 1997.

16. Katz JF: Recurrent Legg-Calve-Perthes disease. JBone Joint Surg 55A:833–836, 1973.

17. Kemp HBS, Chomeley JA, Baijens JK: RecurrentPerthes disease. Br J Radiol 44:675–681, 1971.

18. Khermosh O, Wientroub S: Dysplasia epiphysealiscapitis femoris. J Bone Joint Surg 73B:621–625,1991.

19. Kleinman RG, Bleck EE: Increased blood viscosityin patients with Legg-Perthes disease: A preliminaryreport. J Pediatr Orthop 1:131–135, 1981.

20. Legg AT: An obscure affection of the hip-joint.Boston Med Surg J 162:202–204, 1910.

21. Martinez AG, Weinstein SL: Recurrent Legg-Calve-Perthes disease. Case report and review of the litera-ture. J Bone Joint Surg 73A:1081–1085, 1991.

22. Meyer J: Dysplasia epiphysealis capitis femoris.Acta Orthop Scand 34:183–197, 1964.

23. Pedersen EK: Dysplasia epiphysialis capitis femoris.J Bone Joint Surg 42B:663, 1960.

24. Perthes G: Ueber arthritis deformans juvenilis. DtschZeitschr Chir 107:111–115, 1910.

25. Sanchis M, Valencia S, Zahir SA, et al: The experi-mental simulation of Perthes disease by consecutiveinterruptions of the blood supply to the capital fem-oral epiphysis in the puppy. J Bone Joint Surg55A:335–342, 1973.

26. Skaggs DL, Tolo VT: Legg-Calve-Perthes disease. JAm Acad Orthop Surg 4:9–16, 1996.

27. Waldenstrom H: Der obere tuberkulose collumherd.Z Orthop Chir 24:81–92, 1909.