Pictorial Essay MR Imaging in Juvenile Rheumatoid Arthritis

Document Sample
Pictorial Essay MR Imaging in Juvenile Rheumatoid Arthritis Powered By Docstoc

                                                                                                                                                                                                                Pictorial Essay

‘:                           . .       .   .     .     .    . . . .,       :
                                                                           ::   ,   ‘   #{149}                    ,

     MR Imaging                                                        in Juvenile                              Rheumatoid                                         Arthritis
     Melvin 0. Senac, Jr.,1 Drew                                               Deutsch,2           Bram        H. Bernstein,3            Philip        Stanley,1                 John              V. Crues                 111,24 David              W. Stoller,5
     and Jerrold Mink2

           Juvenile         rheumatoid                     arthritis           is a chronic         inflammatory               dis-     dence of disease, and 12 of the 15 hips showed clinical evidence of
     ease that affects approximately  100,000 children in the United                                                                    arthritis.
     States. Although     the disease can affect multiple organs,                                                                          All scans were performed on a 1 .5-T MR scanner (General Electric
                                                                                                                                         Medical        Systems,                Milwaukee,               WI).     Fifteen          hips     in eight      patients          were
     inflammatory   joint changes are its most prominent manifes-
                                                                                                                                        imaged          in the      coronal                plane    by using             a quadrature                body    coil.      A spin-
     tations. Because of the abundance               of nonossified       growth and
                                                                                                                                        echo       sequence                of        500-or-2000/20-and/or-60                                 (TR/TE)          was          used.
     articular cartilage    in children,     conventional        radiography     is not                                                 Seven of these hips subsequently                                          were imaged in the sagittal plane
     a sensitive   means of assessing           subtle cartilaginous         and soft-                                                  with 2000/20,               60. All but two                   of these patients                     were imaged              by using
     tissue changes      in the early stages of the disease.                                                                            a 256 x 128 data-acquisition       matrix. The remaining two patients
         MR imaging offers the ability to study disorders                   of periph-                                                  were imaged by using a 256 x 256 acquisition matrix. All scans were
     era] joints noninvasively        [1 2]. The joints of 21 patients with
                                                                       ,                                                                performed    with two excitations.
     clinically proved juvenile rheumatoid              arthritis were evaluated                                                            Twenty   knees were evaluated by MR in 1 9 arthritic patients. The
     to determine     the spectrum       of abnormalities         seen with MR and                                                      first five patients were imaged by using a 12.1 -cm, rounded surface
     to compare      these findings       with those of conventional            radio-                                                  coil    placed          posterior              to     the     knee.           The     subsequent               1 5 knees             were
                                                                                                                                        imaged by using a specially designed                                                send-and-receive                 circumferen-
     graphs. In addition,      joints in three volunteers           were studied to
                                                                                                                                        tial surface            coil (General                  Electric         Medical         Systems,              Milwaukee,              WI).
     obtain a normal baseline.
                                                                                                                                        Two excitations                  were used with the flat surface coil,                                         and     one excita-
                                                                                                                                        tion    was       used       with            the     extremity           coil.       Images          were      obtained             in the
                                                                                                                                        coronal         plane       with         1 000/20.            This was followed                      by a sagittal             acquisi-
     Materials              and Methods
                                                                                                                                        tion    by using           2000/20,             60. A field of view of 12-1 6 cm and a data-
       The hips and knees of three normal volunteers (ages: 7, 1 1 and                                                     ,            acquisition             matrix          of    256 x 128 were used in all knees (except for
     15 years, respectively) and 21 patients who had clinically proved                                                                  the     last     four      knees             studied,         for       which        a matrix          of 256         x      256      was
     juvenile       rheumatoid                 arthritis      were         studied.       The    mean    age   of the group             used). All scans were performed with a 3- to 5-mm slice thickness
     with juvenile rheumatoid arthritis was 1 1 .2 years (range, 6-20 years),                                                           with no interslice spacing.
     and the mean duration of disease was 5.4 years (range, 2-8 years).                                                                     Anteroposterior  radiographs of the hips and anteroposterior   and
     Two patients had follow-up scans. No sedation was used, nor were                                                                   lateral radiographs of the knees were obtained in each child within 2
     any       other    joints        studied.             All the         knees        imaged     showed       clinical       evi-     weeks          of the MR study.                       The MR images                    in each        case      were         reviewed

        Received         September    23, 1987; accepted    after revision December  15, 1987.
        Presented         at the inaugural  conjoint meeting of the Society for Pediatric Radiology                                       and the European                  Society           of Pediatric            Radiology,          Toronto,      Ontario,       Canada,
     June 1987.
           1   Department          of Radiology,            Childrens Hospital of Los Angeles, 4650 Sunset Blvd.,                         Los Angeles,             CA 90027.                Address         reprint      requests         to M. 0. Senac,             Jr.
           2   Department          of Radiology,            Cedars Sinai Medical Center, Los Angeles, CA 90048.
           3   Department   of Rheumatology,       Childrens    Hospital of Los Angeles,   Los Angeles,   CA 90027.
           4   Present address:   Department      of Radiology,      Santa Barbara Cottage   Hospital,  Santa Barbara,                                          CA 93102.
           5   Department   of Radiology,    University    of California  Medical Center, San Francisco, CA 94143.
     AJR 150:873-878,                 April          1988 0361-803x/88/1504-0873                        C American     Roentgen       Ray Society
874                                                                                         SENAC         ET AL.                                                                         AJA:150,    April 1988

   Fig. 1.-MR      Images of normal hips (SE 500/20).
   A, Coronal    Image in a 7-year-old  boy shows generous                      epiphyseal,    articular,    and acetabular cartilage          (arrows).
   B, Coronal    image shows thinner   generalized  cartilage                   In this 11-year-old       female volunteer,        reflecting  less epiphyseal     growth   cartilage.
   C, Sagittal   image (SE 2000/20)          in same 11-year-old            girl as B shows better definition of hyaline articular cartilage                   than coronal   image       does.     Separation
of acetabular     (small   arrow)   from   femoral      head    cartilage     (large arrow) Is seen. In general,            cartilage      was better seen In sagittal plane.

   Fig. 2.-Advanced       juvenile rheumatoid     arthritis of hips in a 9-year-old   girl.
   A, Conventional    radiograph   shows marked joint-space          narrowing and femoral head erosions.
   B, Coronal    MR Image (SE 1000/20)        shows complete      loss of articular and epiphyseal cartilage,                        flattening   of left femoral head, and fluid            in subchondral
cyst of right femoral head (arrow).
   C, More anterior coronal Image shows pannus invading subchondral                 cyst (arrow).

by two    radiologists      and     were   correlated          with   corresponding         radio-          epiphysis   best, and cartilage was most clearly identified in the
graphs.                                                                                                     sagittal plane (Fig. iC).
                                                                                                               Of the 1 5 hips examined,      three were interpreted  as normal
Results                                                                                                     on conventional     radiographs       and two were interpreted    as
                                                                                                            normal on MR. Cartilage        loss was seen in 1 3 of 1 5 hips on
    In every patient,      diagnostic     MR images           were obtained,                                MR and was indirectly suggested by joint-space narrowing in
although     in one patient a minor degree of degradation                 of the                            1 2 of 15 hips on standard films. In eight hips, cartilage loss
images occurred        because of motion.                                                                   as shown        by MR appeared          to be more severe than was
    Normal       hips showed       smooth      articular     and epiphyseal                                 suggested        by conventional       studies    (Figs. 2 and 3). Joint
growth     cartilage  of intermediate      signal intensity       surrounding                               effusions    were detected       in six hips, one of which was radio-
the femoral head; the growth cartilage                was easily separated                                  graphically     visible. Avascular       necrosis    of the femoral     head
from the ossified cortex (low intensity)              and the high-intensity                                was seen in two of 1 5 hips with equivocal plain film findings.
epiphysis     (Fig. 1). Cartilage     became       thinner with increasing                                  An intraarticular      fragment   was seen in one hip on MR without
age (Fig. 1 B). There were no joint effusions              in the normal hip.                               radiographic       abnormality.    Bone erosions        and secondary     de-
Coronal     images showed         the integrity      of the capital femoral                                 generative      changes      were seen by both techniques             in nine
AJR:150, April 1988                                           JUVENILE            RHEUMATOID          ARTHRITIS                                                               875

    Fig. 3-Cllnlcally    symptomatic   left hip in an 1 1-year-old girl with juvenile rheumatoid   arthritis.
    A, Conventional radiograph shows no appreciable joint space narrowing or evidence of hip involvement
    B, Coronal     image (SE 500/20) shows generalized thinning of cartilage without bone erosions. Also note closing growth plate on radiograph, difficult to
identify  on MR.
    C, Coronal image (SE 500/20)      in 11-year-old    female volunteer    shows normal articular    and residual epiphyseal cartilage for this age group (arrows).
Compare     with B.

     Fig. 4.-Normal       knee in a 7-year-old girl.
     A, Coronal      image (SE 1000/20) shows uniform high signal intensity        of epiphyses         and intermediate    signal cartilage.   Epiphyseal (growth)   and hyaline
articular    cartilage   are difficuft to distinguish from each other. Low-signal   posterior      cruciate ligament is seen along inner surface of medial femoral condyle
     B, Sagfttal image (SE 2000/20) shows thick cartilage and bow tie-shaped              menisci.
     C, On this sagfttal     image, note smooth and slightly concave     posterior surface of infrapatellar        fat pad (arrows),   a site of normal synovial reflection.  Also
note patellar cartilage easily separable from femoral cartilage.

hips, but were generally         more extensive      on MR.                                    The infrapatellar   fat pad was of high signal intensity,             with a
    Follow-up     studies   showed     progressive      articular  cartilage                   smooth and slightly concave        posterior    surface. Separation          of
loss with increasing      joint fluid in the hip of one child who was                          articular cartilage from growth cartilage was not possible, but
more symptomatic           clinically  but whose       radiographs       were                  cartilage in general was of uniform thickness,            becoming      grad-
unchanged.                                                                                     ually thinner with age (Fig. 5). Hyaline and epiphyseal            cartilage
    In the normal knee (Fig. 4), articular and epiphyseal            growth                    was best seen on sagittal images. There were no joint effu-
cartilage     has a homogeneous,          intermediate      signal intensity                   sions, focal cartilaginous     changes,      or bone erosions          in the
greater    than the adjacent           low signal       intensity        cortex     and        normal knee.
fibrocartilaginous  menisci.          Normal      menisci were       seen as tn-                   Of the 20 knees examined,        MR showed          articular  and epi-
angles of decreased    signal         intensity    on all pulsing    sequences.                physeal cartilage   loss in 1 4, focal thinning      in five, and uniform
876                                                                              SENAC      ET AL.                                                          AJA:150,   April 1988

    FIg.    5.-Normal        knee In a 15-year-old boy.             Fig. 6.-Advanced     juvenile  rheumatoid arthritis In a 9-year-old girl.
 Sagittal  image(SE        2000/20)shows       articular hya-       A, Anteroposterior radiograph shows epiphyseal overgrowth and joint-space narrowing.
 line cartilage (arrowheads).         Apparent     thinning of      B, Sagittal   MR image (SE 2000/20) shows severe loss of articular and epiphyseal cartilage and
 cartilage   in this age group reflects         less epiphy-     marked     degeneration   of menisci.  Hypoplasia or near-complete destruction of menisci was a
 seal (growth)     cartilage.                                    common     MR finding in juvenile rheumatoid arthritis.

    Fig. 7.-Minimally    symptomatic   right knee in a 10-year-old      girl with juvenile rheumatoid   arthritis.
    A, Anteroposterior    radiograph shows osteopenia     and mild epiphyseal         overgrowth.
    B, Sagittal   image (SE 2000/60) shows small menisci       (arrows)      and diffuse thinning of cartilage,    particularly  undersurface     of femoral condyle and tibial
plateau (arrowheads).
    C, Slightly more medial image shows small knee effusion            (arrow)    and mild irregularity of posterior     surface of infrapatellar    pad Indicative of synovltis

loss in nine (Fig. 6). Cartilage         loss, suggested           indirectly     on          was of intermediate  signal intensity on Ti -weighted  images
radiographs    by joint-space     narrowing,        was seen in nine of 20                    and was seen as high signal intensity on T2-weighted  images.
knees. In 1 3 of 20 knees, meniscal changes were present.                          In         MR showed popliteal cysts in i 0 knees; only one was shown
six, the menisci     were completely            atrophic,      whereas       in the           radiographically     (Fig. 8C). Other changes            in juvenile     rheuma-
other seven, they were hypoplastic               with shortening         and loss             toid arthritis    shown on MR (Fig. 8) were avascular         necrosis
of the normal triangular      appearance        (Fig. 7). Inflammation         was            (one), intraarticular    fragments    (three), and medullary    infarcts
observed    at sites of synovial reflection          in 1 5 knees. This was                   (one), none of which was visible on corresponding                  radio-
best seen along the posterior,           concave        surface of the infra-                 graphs.
patellar fat pad, which became             irregular      with synovitis       (Fig.              Follow-up      MR studies     were obtained     in one knee that
7C). Fifteen knees showed          joint effusions         that were seen in                  showed       progressive    avascular   necrosis  of the femoral con-
 nine knees on conventional           radiographs         (Fig. 8). Joint fluid               dyles and tibial plateau.
AJR:150,        April 1988                                              JUVENILE   RHEUMATOID          ARTHRITIS                                          877

      FIg. 8.-Other     findings In juvenile     rheumatoid    arthritis.
      A, Radiograph shows large joint effusIon            In an 8-year-old boy.
      B, Sagittal   Image (SE 2000/60)       In same child as in A shows not only joint effusion      but also   an osteochondral   fragment   (arrow).
      C, Sagittal Image (SE 2000/20)         shows popliteal     cyst (arrow) in a 15-year-old boy.

    Fig. 9.-Moderately             symptomatic     left knee in
12-year-old    girl with        juvenile  rheumatoid      arthri-
    A and         B, Ti-weighted         saglttal   images       (SE
 1000/20)        show MR findings of hypoplastic                me-
nlsci and        irregularity    of Infrapatellar    fat pad     (ar-
row), but        cartilage    appears    reasonably      well   pre-
     C and D, 12-weighted     Images (SE 2000/60)
show that most of this apparent     cartilage is ac-
tually joint effusion  (long arrows in D) with high
signal       Intensity   as   opposed     to thinned     articular
and residual eplphysealcartllage        of intermediate
Intensity (short arrows       In C). T2-weighted        im-
ages are necessarytodifferentlateeffusion            from
878                                                                              SENAC    ET AL.                                                                  AJR:i50.    Apr    1988

                                                                                                                                Fig.                   juvenile rheumatoid
                                                                                                                                    In a 20-year-old woman with exacerba-
                                                                                                                           tion of symptoms In left knee clinically suspi-
                                                                                                                           cious for avascular necrosis.
                                                                                                                               A, Anteroposterior         radiograph       shows joint-
                                                                                                                           space      narrowing       of medIal     compartment       and
                                                                                                                           nonspecific      subchondral      sclerosis.
                                                                                                                               B, Coronal       Image (SE 500/20)         shows exten-
                                                                                                                           sive avascular        necrosis  of tiblal plateau    and me-
                                                                                                                           dial femoral condyle. Also note medullary                    in-
                                                                                                                           farcts    in metaphysis        of tibia.     Decompression
                                                                                                                           core biopsies        confirmed    diagnosis     of avascular

Discussion                                                                                 partial explanation   for thejoint   instability so frequently   present
                                                                                           in children   with juvenile     rheumatoid      arthritis. The cause of
    Juvenile    rheumatoid      arthritis  is a chronic, generalized     col-
                                                                                           these meniscal changes         is not known, but it may be related
lagen vascular        disease with variable multisystem         manifesta-
                                                                                           to joint effusions   associated      with the release of proteolytic
tions. The age of onset is variable, but approximately               80% of
                                                                                           enzymes       that can lead to the degeneration                        of cartilaginous
children     become      symptomatic       by the time they are 7 years
                                                                                           structures   [6].
old [3, 4]. The hips and knees frequently are affected, but the
                                                                                              Avascular     necrosis was seen in two hips and in one knee
radiographic       features    of joint-space     narrowing   and erosive
                                                                                           in our series. All three patients    had severe joint pain but
bone changes usually do not occur until 2 or more years after
                                                                                           showed       no radiographic            evidence     of this entity.        In one child,
the onset of disease [4J. The abundance of nonossified car-
                                                                                           progression      of necrosis     was seen in the knee on follow-up
tilage is the major reason for the relatively late radiographic
                                                                                           imaging. Despite the striking MR changes,               radionuclide     bone
                                                                                           scanning     and standard      radiographs      failed to show evidence
  In our study, MR detected          fluid within the hips and knees
                                                                                           of avascular     necrosis (Fig. i 0). Decompression           core biopsies
much more frequently    than         conventional          radiographs       did.
                                                                                           confirmed     nonviable    bone.
Whereas hip effusions have been detected                   by MR in 84% of
                                                                                               In summary,      this investigation     of 2i patients     with juvenile
normal adults [5], these were not seen in any of our normal
                                                                                           rheumatoid      arthritis indicates     that MR can play a significant
hips. Joint fluid was well visualized             on both Ti          and T2-
                                                                                           clinical role in the evaluation          of rheumatic     joint disease       in
weighted     images. On Ti -weighted           images,      however,     it was
                                                                                           children.    The extent and detail of the changes                 of juvenile
difficult to separate effusions     from articular cartilage because
                                                                                           rheumatoid      arthritis   were seen better with MA than with
of their similar signal intensities       (Fig. 9). Therefore,        in evalu-
                                                                                           conventional      radiographs,       and progression      of disease      was
ating children with juvenile rheumatoid            arthritis,    T2-weighted
                                                                                           also seen. Clinical studies of children 6 years old and younger
images are recommended      because on T2-weighted                             images
                                                                                           who may require sedation                     seem warranted,          as does evalua-
fluid appears higher in signal intensity than does                            adjacent
                                                                                           tion of additional      smaller joints commonly    affected in juvenile
                                                                                           rheumatoid     arthritis.     MA can be used to detect active inflam-
   Once the normal MR appearance                   ofgrowing       hips and knees          matory changes          within the joint and thus may be a valuable
in children     had     been   determined,         it was    possible        to detect
                                                                                           diagnostic   technique        to monitor and perhaps alter treatment.
subtle cartilaginous and synovial changes in a child with
juvenile rheumatoid  arthritis. Cartilage loss was present in
most knees and hips, generally appearing worse than the                                    REFERENCES
corresponding         radiographs     suggested.        Early     synovial      inflam-     1 . Totty WG, Murphy WA, Ganz WI, Kumar B, Daum WJ, Siegel B. Magnetic
mation was       detected    on MR by observing      irregular borders                           resonance    imaging     of the normal      and ischemic         fernoral   head. AiR
                                                                                                 1984;143: 1273-1280
of synovium      at their sites of reflection.  More advanced     syno-
                                                                                            2. Aeicher MA, Hartzman 5, Bassett LW, Mandelbaum                    B, Duckwiler    G, Gold
vial changes      showed    soft-tissue   masses of intermediate    sig-
                                                                                                 RH. MR imaging of the knee. Radiology 1987:162:547-551
nal within the joint surrounded    by effusion of high intensity.                           3. Coss JA, Boots RH. Juvenile rheumatoid            arthritis; study of fifty-six cases
The demonstration    of hyperplastic   synovium by MA, which                                     with note on skeletal changes. J Pediatr 1946;29:143-i56
was   not seen        on standard    radiographs,           may   have       important      4. Martel W, Holt JF, Cassidy JT. Roentgenologic               manifestations     of juvenile
                                                                                                 rheumatoid   arthritis. AiR 1962:88:400-423
clinical implications in terms of suggesting                 the need for more
                                                                                            5. Mitchell DG, Rao V, Dalinka M, et al. MRI of joint fluid in the normal and
aggressive medical or surgical therapy.                                                          ischemic hip. AJR 1986;146:1215-1218
     Atrophic or hypoplastic          menisci were detected in the ma-                      6. Harris ED. Role of collagenase       in joint destruction.      In: Sokoloff   L, ed. The
jority of knees evaluated           by MR. This may provide at least a                          joints and synovial fluid. New York: Academic            Press, 1978:243-266