FATAL DESTRUCTIVE CERVICAL SPONDYLOARTHROPATHY IN TWO PATIENTS ON LONG TERM DIALYSIS
Jeffrey C. Allard
Maria E. Artze
Guy Porter
G. O. Perez
L. Ghandur-Mnaymneh
R. deVelasc
From the Departments of Radiology, Medicine, and
Pathology, University of Miami School of Medicine, and the
Veterans Affairs Medical Center, Miami, Florida.
Correspondence to JCA, PO Box 016960 ® 109), Miami, FL
33101, Tel. 305 549 7844
FATAL AZOTEMIC DESTRUCTIVE SPONDYLOARTHROPATHY; FINDINGS- -AND
IMPLICATIONS
Abstract
Two patients with fatal cervical cord compressive
myelopathy are described, both of whom had been on
dialysis for over fifteen years. Destructive changes were
noted in mid and upper cervical regions, with soft tissue
mass in atlanto_occipital region in one patient. Clinical,
radiographic, and biopsy findings suggest both amyloid and
hyperparathyroidism as etiologies for these destructive
spinal changes. These cases emphasize the aggressive
potential of destructive spondyloarthropathy associated
with dialysis which has not been reported to date.
Introduction
Erosive arthropathy in patients on chronic dialysis
was initially described in the hands. Because up to 30% of
patients had these findings and were asymptomatic, the
entity was initially felt to be "innocuous" [1]. Since
1980 the spectrum and severity of changes in renal
osteodystrophy has been expanded to include erosions in
the axial as well as the peripheral skeleton [2]. A
destructive spondyloarthropathy (DSA) associated with
dialysis was first described in 1984, with a total of 73
cases being reported to date [3-15]. Although
approximately 20% of ADS cases have had neurologic
compressive symptoms, there have been no cases of
paralysis or death attributed to the spinal lesions thus
far [2-15]. We report two cases of fatal cervical cord
compression related to ADS and discuss etiology and
clinical implications.
Case reports
Case one
A 51_year_old woman on maintenance hemodialysis for
17 years was admitted for neck pain and quadriparesis.
Imaging studies disclosed odontoid destruction with
associated mass which caused cervical cord compression as
well as peripheral arthropathy and other evidence of renal
osteodystrophy (Fig. 1). Surgical specimen revealed
fibrocartilage with degenerative change and synovial
hyperplasia with hemosiderin and fibrosis. No amyloid or
crystals were found, despite electron microscopy. Patient
expired two months later with evidence of progressive
neurologic deterioration
Previous history included multiple blood
transfusions, spontaneous rib fractures, hip and knee
joint replacement surgery, and carpal tunnel surgery.
Laboratory values included negative rheumatoid factor,
normal aluminum levels, and calcium, phosphorus, and
alkaline phosphatase levels which varied from normal to
elevated.
Case two
A 67 year old man on dialysis for 15 years was
admitted for neck pain and spastic paraplegia. Seven years
previously he had undergone anterior decompressive
diskectomy with fusion at C3-5 for spinal stenosis and
cervical root syndrome. Imaging studies at that time
demonstrated findings of DSA involving C3-4 and C5-6
levels (Fig.2a)
On this admission, MRI disclosed increased kyphosis
from C3 to C6 with spinal canal stenosis and cord atrophy
(Fig. 2b). Erosive arthropathy of hands with carpal cyst
formation was also present and noted to have progressed
from previous studies (Fig. 2c). Iliac crest bone biopsy
showed aluminum deposits and increased osteoclastic and
osteoblastic activity consistent with secondary
hyperparathyroidism. There was marked elevation of serum
parathormone, calcium, and phosphorus levels. Patient
expired prior to surgical decompression related to sepsis
from complicating decubitus ulcers.
Discussion
Although the frequency of spinal erosive changes
associated with dialysis is up to 25%, destructive changes
are less common [6]. The frequency of changes increases
with the duration of dialysis and a recent report also
indicates that dialysis is not required for the
development of DSA, but that duration of renal failure may
be the more important factor [4,13]. Most cases of DSA
with follow-up have demonstrated relatively rapid
progression of destructive changes over a period of months
[2-12]. Early radiologic findings include erosions of
anterosuperior or anteroinferior corners of vertebral
bodies which are similar to the Romanus lesion in
ankylosing spondylitis [6,9]. Osteophytes are absent.
Later changes include disk space narrowing with
associated end plate erosions and irregularity caused by
cystic erosions and end_plate sclerosis [4,8]. CT
scanning, as in our cases, best demonstrates the erosions
which are well defined and variably sized [5-8,14]. At
this stage differentiation from infection or neuropathic
change is difficult with plain films alone. Needle biopsy,
bone scan, and MRI have been used to exclude infection in
some cases [3,6-10]. Both of our cases had MRI and showed
no evidence of T2 prolongation to indicate inflammation or
infection, in agreement with other reported cases except for
one [9]. Erosions at multiple sites in the appropriate
clinical context, as in our cases, indicates that
infection is less likely [3-13]. In addition, the absence
of associated paravertebral swelling or soft tissue mass
has ruled against infection [6,8,10].
Most reported cases of DSA have involved the cervical
spine, and multilevel disease is common [4-14]. Mid¬cervical
disease is most frequent. Lumbar is more common
than thoracic spine involvement. Vertebral subluxation is
an increasingly recognized complication of DSA and is of
particular importance because of potential neurologic
complications. 13 cases of DSA in the literature have had
subluxations, particularly at the C3-4 level [5-8,10-13,15].
Of five reported instances of surgical
intervention, all cases had subluxation. Three cervical,
one thoracic, and one lumbar decompression have been
performed [5,10,11,13]. Three additional cases of cord
compression have been mentioned which did not require
surgery [5,14]. Spinal subluxations have also been noted
in dialysis patients independent from DSA, suggesting
possible ligamentous laxity in these patients [15].
A recent report has described pseudotumors of the
craniocervical junction on MRI, defined as retrodental
soft tissue wider than 5 mm, in 28% of a population on
dialysis for greater than 10 years [14]. Prior to this
report, only one other case of atlantoaxial involvement
had been reported which had unilateral narrowing [9]. Only
one of the pseudotumor patients had plain film findings.
In 5 of 7 pseudotumor patients, CT demonstrated bony
cystic erosions adjacent to the soft tissue mass [14]. The
absence of other rheumatoid features such as subluxation
was noted in all cases. Our case one is an advanced
example of pseudotumor formation with associated
horizontal atlanto_axial subluxation. As with
intervertebral imaging in DSA, our pseudotumor like others
reported was low to intermediated signal on T2-weighted
MRI. One should also appreciate that pathological and
radiologic changes at atlanto-axial level are likely to be
unique as this is a synovial joint.Many etiologic agents have been postulated in ADS,
including amyloid, crystal deposition, iron, and aluminum
[2_16]. Only biopsy specimens containing amyloid (beta 2
microglobulin){n=10}, hydroxyapatite {n=2}, and
pyrophosphate crystals {n=1} have been documented to date
[3_5,7,9,11,14]. We were able to demonstrate neither
crystals nor amyloid in our biopsy specimens. The
discovery of aluminum in one of our specimens is probably
not related to DSA Indeed, aluminum deposits are present
in up to half of all bone biopsy specimens in chronic
dialysis patients and are correlated only with the
presence of osteomalacia [2,17].
Although some initial reports suggested crystal
deposition as etiologic, recently reported cases have
failed to show these [2-13]. Generally, chondrocalcinosis
is probably no more frequent in dialysis patients than in
age_matched controls, which should also discredit the
crystal theory [4]
Recovery of beta 2 microglobulin, the specific
amyloid protein which is elevated in dialysis patients, is
more frequent. In several cases, however, recovery of
amyloid has been distant to the site of DSA, particularly
in the cases with carpal tunnel syndrome [4,14].
Peripheral juxtarticular cystic changes have also been
demonstrated to contain amyloid, particularly when located
in the carpal bones or heads of humerus or femur [16].
Some authors have used this association as collaborative
evidence for an amyloid etiology in DSA [5,9,14,15]. That
one of our cases had carpal tunnel surgery and both had
carpal cysts is our only evidence for an amyloid etiology.
Although discal, synovial, and perivertebral accumulations
of amyloid have been found, it is still unclear as to how
this results in DSA [2,3,5].
Severe hyperparathyroidism has been noted in
approximately 80% of DSA patients, many of whom underwent
parathyroidectomy [3-5,12-14]. In at least three patients
with DSA, surgical parathyroidectomy resulted in dramatic
symptom relief, and a stabilization of radiographic
findings [12]. Peripheral erosive arthropathy has correlated
poorly with hyperparathyroidism in most cases
and correlated variably with DSA [2-7,9]. Correlation of
subperiosteal erosions and rugger_jersey spine has not
been studied frequently enough to make conclusions. Both
of our patients had biochemical evidence of secondary
hyperparathyroidism. In addition, case one had central
osteosclerosis and case two had biopsy evidence to support
this as an etiology. Although parathormone might cause
bony resorption with associated ligamentous weakening, the
exact mechanism of DSA production remains speculative, and
perhaps is multifactorial.
In summary, DSA is an increasingly recognized
and sometimes lethal process associated with long standing
renal failure usually in conjunction with dialysis.
Radiographically, the presence of disk space narrowing
with irregular end plate destruction resembles infection
and biopsy may be needed to exclude this diagnosis when it
is clinically suspected. Pseudotumor formation in
atlantoaxial region, with bony erosion, is a second
manifestation of ADS. As ADS is most frequent and
potentially lethal in the cervical region, we recommend
lateral radiographs and perhaps MRI of this region
especially if the patient is symptomatic, or has been on
hemodialysis for more than five years
References
1. Sundaram M, Wolverson MK, Heiberg E, Grider RD. Erosive
azotemic osteodystrophy. AJR 1981;136:363-367
2 Sundaram M. Renal osteodystrophy. Skeletal Radiol
1989;18:415-426
3. Kuntz D, Naveau B, Bardin T, Drueke T, Treves R, Dryll
A. Destructive spondyloarthropathy in hemodialyzed
patients:a new syndrome. Arthritis Rheum 1984;27:369-375
4. Bindi P, Chanard J. Destructive spondyloarthropathy in
dialysis patients: an overview. Nephron 1990;55:104-109
5. Orzincolo C, Bedani PL, Scutellari PN, et al.
Destructive spondyloarthropathy and radiographic follow_up
in hemodialysis patients. Skel Radiol 1990;19:483-487
6. Sundaram M, Seelig R, Pohl D. Vertebral erosions in
patients undergoing maintenance hemodialysis for chronic
renal failure. AJR 1987;149:323-327
7. Kaplan P, Resnick D, Murphy M, Heck L, Phalen J, Egan
D, Rutsky E. Destructive noninfectious spondyloarthropathy
in hemodialysis patients: a report of four cases.
Radiology 1987;162:241-244
8. Rafto SE, Dalinka MK, Scheibler ML, Burk DL, Kricum ME.
Spondyloarthropathy of the cervical spine in long-term
hemodialysis. Radiology 1988;166:201-204
9. Naidich JB, Mossey RT, McHeffrey-Atkinson B, et al.
Spondyloarthropathy from long_term hemodialysis. Radiology
1988;167:761-764
10. Patel B, Mistry CD, Kumar EN, Short CD, Jenkins JPR.
Magnetic resonance imaging in non_infective destructive
spondyloarthropathy. Br J Radiol 1988;61:511-514
11. Sebert JL, Fardellone P, Marie A, et al. Destructive
spondyloarthropathy in hemodialyzed patients: possible
role of amyloidosis. Arthritis Rheum 1986;29:301-302
12. McCarthy JT, Dahlberg PJ, Kreigshauser JS, et al.
Erosive spondyloarthropathy in long-term dialysis
patients: relationship to severe hyperparathyroidism. Mayo
Clin Proc 1988; 63:446-452
13. Alcalay M, Goupy MC, Azais I, Bontoux D. Hemodialysis
is not essential for the development of destructive
spondyloarthropathy in patients with chronic renal
failure. Arthritis Rheum 1987;30:1182-1186
14. Rousselin B, Helenon O, Zingraff J, et al. Pseudotumor
of the craniocervical junction during long-term
hemodialysis. Arthritis Rheum 1990;33:1567-1571
15. Hardin P, Flipo RM, FoissocGegoux P, et al. Current
aspects of osteoarticular pathology in patients undergoing
hemodialysis:study of 80 patients. J. Rheum. 1987;14:780¬783
16. Sargent MA, Fleming SJ, Chattopadhyay C, Ackrill P,
Sambrook P. Bone cysts and haemodialysis_related
amyloidosis. Clin Radiol 1989;40:277-281
17. Berland Y, Grandvuillemin M, Charhon S, Olmer M.
Aluminum and renal osteodystrophy in chronic uremia. KidIntl 1985;28:S183-S187
Figure legends
_1a. Sagittal SE 1800/60 one year prior to admission shows
pseudotumor (white arrows) surrounding odontoid and
narrowing spinal canal.
1b. Post myelography CT at same time shows multiple cystic
erosions of odontoid and C2 with large retrodental soft
tissue mass which impinges on anterior subarachnoid space.
These changes progressed on studies one year later
resulting in cord compression
1c. Advanced erosive arthropathy of hands involving both
distal and proximal portions. Cortical tunneling indicates
secondary hyperparathyroidism.
1d. Central osteosclerosis, healed right rib fractures,
and advanced erosions of left humeral head shown1e. Marked joint space narrowing with irregular erosion of
medial tibial plateau led to joint replacement.
1f. Lateral operative view confirms needle placement in
region of pseudotumor. There is significant anterior
atlanto_axial subluxation.
2a. Lateral cervical view 7 years prior to admission shows
disk space narrowing at C3-4 and C5-6 with end plate
erosions, irregularity, and sclerosis which is
characteristic for DSA.
2b. CT at same time at C5 level shows cystic erosions in vertebral body and lateral mass much better than plain
film.
2c. SE 600/20 sagittal MR at time of admission shows
cervical malalignment with kyphosis and cord compression
with atrophy from C3 to C6. Note metallic artifact
anteriorly at C5-6 level related to previous surgery.
2d. Hands demonstrate cortical tunneling, juxtarticular
erosions, and carpal cysts.