Arch Orthop Trauma Surg (2002) 122: 73–79 DOI 10.1007/s004020100321
O R I G I N A L A RT I C L E
Masahiro Inaoka · Koichi Tada · Kazuo Yonenobu
Problems of posterior lumbar interbody fusion (PLIF) for the rheumatoid spondylitis of the lumbar spine
Received: 26 January 2001 / Published online: 13 September 2001 © Springer-Verlag 2001
Abstract We performed posterior lumbar interbody fusion (PLIF) on 7 patients with rheumatoid spondylitis (RA) of the lumbar spine with severe low-back pain and/or cauda equina symptoms, and evaluated the effectiveness of PLIF for the lumbar spinal instability in RA secondary to destruction of the anterior elements, including vertebral endplates and the apophyseal joint. The subjects were 7 patients with classic RA, 2 men and 5 women, mean age 65 years old, and the mean duration of RA was 21 years. All had severe low-back pain and difficulty with walking. According to the ARA classification, the patients were at stage 3 or worse and in class 3. Diagnostic imaging including magnetic resonance imaging (MRI), tomography, myelography, and computed tomographic myelography (CTM) of the lumbar spine clearly delineated pathology, destruction of the vertebral endplate, subluxation, and cauda compression which can be well treated with PLIF. We performed L4/5PLIF (5 cases), L3/4 and L4/5 PLIF (2 cases), and posterior fixation with instruments for anterior column repair and stabilization and posterior decompression. Autografts (all cases) and Brantigan IF cage (2 cases) were used. Stable fixation of the lumbar spine was achieved after surgery, and improvement in gait and activities of daily living were achieved through the relief of low-back pain and radicular pain; the mean duration of follow-up was 22 months. Postoperative, plain radiography, CT, and MRI revealed the enlargement of the lumbar canal and fusion and incorporation of grafted bone, but in some cases, collapse of graft, migration of pedicle screw,
M. Inaoka (✉) Department of Orthopaedic Surgery, Nissay Hospital, 3–8, Itachibori 6-chome, Nishi-ku, Osaka 550-0012, Japan e-mail:
[email protected], Tel.: +81-6-6543-3581, Fax: +81-6-6532-6482 K. Tada Department of Orthopaedic Surgery, Kansai Rosai Hospital, 3–1-69, Inabaso, Amagasaki, Hyougo 660, Japan K. Yonenobu Department of Orthopaedic Surgery, Osaka University Medical School, Suita, Osaka, Japan
instability of adjacent level, and collapse of adjacent vertebra were noted. PLIF with spinal instruments is a preferred treatment for rheumatoid spondylitis of the lumbar spine, but in the mutilating type of RA with severe osteoporosis, PLIF in combination with a long fixation system and/or augmentation of the vertebral bodies might be needed. Keywords Posterior lumbar interbody fusion · Rheumatoid spondylitis · Rheumatoid arthritis
Introduction There are many reports regarding surgical treatment for spinal lesions associated with rheumatoid arthritis (RA). Most of the reports are related to the cervical spine. Rheumatoid spondylitis of the cervical spine actually causes activities of daily living (ADL) disorders as severe as those associated with arthrosis of the extremities. Courses of cervical treatment based on the type of lesion and stage of morbidity are currently being established. Although in the literature there are several reports on clinical symptoms, imaging findings, pathological findings, and pathology of lesions associated with lumbar spine RA [2, 3, 4, 5, 6, 8, 10, 11, 14], only a few deal with the surgical treatment and prognosis. Involvement of the lumbar spine has been reported in a small number of cases, but the number of patients with RA complaining of lumbar symptoms seems to increase with the aggressive treatment for lower extremity lesions. We performed posterior lumbar interbody fusion (PLIF) on 7 patients with rheumatoid spondylitis (RA) of the lumbar spine with severe low-back pain and/or cauda equina symptoms, and evaluated its effectiveness for the lumbar spinal instability in RA secondary to destruction of the anterior elements, including vertebral endplates and the apophyseal joint. After treatment of the lesions of the lumbar spine, the patients can experience a marked improvement in their level of activity in ADL. However, some complications do occur after the surgery.
74 Table 1 Patients’ data
Case
1 2 3 4 5 6 7
Gender, age (years)
History of RA (years)
Classification
F, 56 F, 65 M, 61 F, 67 F, 74 M, 71 F, 59
7 31 24 9 22 34 23
+ – + – + + +
Level of PLIF (PLF)
Instrument Implant
Follow-up (months)
L4/5, (L4-S) L3/4, L4/5 L4/5 L4/5, L5/S L4/5 L4/5, (L3–S) L4/5, (L3–5)
Diapason Diapason Isola Horizon Diapason Isola Isola
42 43 26 15 13(+) 28 23
Mutilating ARA
Fig. 1a–h Plain roentgenographs of the lumbar spine show mutilation of the L4 vertebral body extending to the caudal endplate (a,b). Both T1- and T2-weighted magnetic resonance (MR) images show a low signal intensity area into the vertebral body at L4/L5 endplate. Severe stenosis of the dural canal was observed at the same level (c,d).Bony union of the grafted bone was observed on plain roentgenography 1 year (e,f) and 3 years after the surgery (g,h)
The purpose of this study was to evaluate the effectiveness of PLIF for the lumbar spinal instability in RA secondary to destruction of the anterior elements, including vertebral endplates and the apophyseal joint. This is a report of the pathology, method of surgery, postoperative results, and problems of the procedure.
4 3 4 3 4 4 4
– AW-GC – I/F cage – AW-GC I/F cage
Patients and methods The subjects were 7 patients with classic RA, 2 men and 5 women, mean age 65 years old, and mean duration of RA 21 years. All had severe low-back pain and difficulty with walking. According to the ARA classification, the patients were at stage 3 or worse and in class 3 (Table 1). All of them complained of severe disruption of ADL due to the exacerbation of low-back pain and neurological symptoms in the lower extremities in spite of conservative treatment. Most major lesions of the lumbar spine were relatively limited to the lower lumbar spine. Bone mutilation was observed as an anterior element from the vertebral endplate into the vertebral body, and osteolysis of the articular process was observed as a posterior element. The common pathology was pain under load in the sitting or standing position, caused by mutilation of bone and cartilage due to RA lesions. This was complicated by cauda equina root symptoms caused by spinal canal stenosis due to vertebral subluxation. We selected PLIF as the common procedure for these 7 patients with
75
Fig. 2a–f Plain roentgenography showed increased sliding of L4 and sliding of the adjacent upper lumbar spine(a,b). Dynamic myelography revealed a blockage at L3 level(c,d). Plain roentgenography 2 years and 3 years, 7 months after surgery showed union of the grafted bone(e,f)
rheumatoid spondylitis of the lumbar spine while we were preparing our surgical treatment plan.
Case reports Case 1 A woman aged 56 years had a 7-year history of RA and was affected with severe mutilation deformities of fingers. She could not remain standing because of severe low-back pain and numbness of the lower extremities. Plain roentgenography (Fig. 1a,b) of the lumbar spine show mutilation of the L4 vertebral body extending to the caudal endplate, especially on the left side, to the anterior endplate. Both T1- and T2-weighted MRI images showed a low signal intensity area into the vertebral body at the L4/5 endplate (Fig. 1c,d). Severe stenosis of the dural canal was observed at the same level. Based on these findings, we selected PLIF, but preservation of the bony endplates was impossible because of the mutilation caused by the RA lesion, and the grafted bone was not expected to open and maintain interbody height. So we used the Diapason system in situ for the fixation. However, since the right L5 pediculus was severely damaged, we inserted pedicle screws into both sides of L4, S1,and in the left side of L5, and performed posterior fixation between L5 and S1. The patient was able to walk in the room and showed marked improvement in ADL at her discharge from the hospital. Bony union of the grafted bone was observed by plain roentgenography 1 year (Fig. 1e,f) and 3 years after the surgery (Fig. 1g,h).
Case 2 A woman aged 65 years old had a 38-year history of RA. She had undergone decompression laminectomy at L4 and 5 for stenosis of the spinal canal as the major pathology. However, low-back pain increased and gait disturbance recurred over 6 years postoperatively. Plain roentgenography showed increased sliding of L4 and sliding of the adjacent upper lumbar spine (Fig. 2a,b). Dynamic myelography revealed a blockage at the L3 level (Fig. 2c,d). In order to treat L3/4 and L4/5 interbody instability, low-back pain, and cauda equina intermittent claudication, we performed double PLIF on L3/4 and L4/5. We used apatite and wollastonite-containing glass ceramics (AW-GC) for the L3/4 interbody bone graft. Plain roentgenography 2 years and 3 years, 7 months after the surgery showed union of the grafted bone (Fig. 2e,f). The patient recovered sufficient gait ability to do housekeeping owing to improvement of the low-back pain and cauda equina symptoms. Case 3 A man aged 61 years had a 23-year history of RA. The patient had no neurological abnormality but was not able to walk because of exacerbation of low-back pain and numbness of the lower extremities in the standing and sitting positions. Plain roentgenography of the lumbar spine showed osteoporosis and anterior sliding of the L4 vertebral body between L4 and 5 (Fig. 3a,b). Tomography of the lumbar spine showed mutilation of the endplate extending to the vertebral body between L4 and 5 (Fig. 3c,d). T2-weighted MR images showed a high signal intensity area into the disk space of L4/5, and thus a differential diagnosis from pyogenic spondylitis
76
Fig. 3a–h Plain roentgenography of the lumbar spine showed osteoporosis and anterior sliding of the L4 vertebral body between L4 and 5 (a,b). Tomography of the lumbar spine shows mutilation of the endplate extending to the vertebral body between L4 and 5 (c,d). Plain roentgenography showed collapse of the grafted bone, increased lateral sliding of L4 body (e,f). Plain roentgenography of the revision of L4/5 by PLIF with pedicle screws into both sides of L2, L3, L5, and S1, and posterior fixation between L2 and S1 (g,h)
Fig. 4a–d Plain roentgenography before the operation (a,b) and 1 year after double PLIF surgery on L3/4 and L4/5, using the Branigan I/F cage for each level (c,d)
was required. CT images revealed the articular processes and vertebral erosion characteristic of RA. Based on these findings, we selected PLIF and PS system in situ for the fixation. The patient was able to walk in the room and showed marked improvement in ADL upon discharge from the hospital. However, his low-back pain increased, and difficulty with walking recurred over 1 year, 6 months postoperatively. Plain roentgenography showed collapse of the grafted bone, increased lateral sliding of the L4 body (Fig. 3e,f), and upper migration of pedicle screws 1 year,7 monthsafter the surgery, leading to revision of L4/5 PLIF. However, since the L4 pedicles and vertebral body were severely damaged, we took out the pedicle screws from L4, inserted pedicle screws into both sides of L2, L3, L5, and S1 (Fig. 3g,h), and performed posterior fixation between L2 and S1. Unfortunately, at 1 year after the surgery, he died of deep infection from sacral decubitus.
77 Table 2 Activities of daily living (ADL) score at preoperation, 3 months postoperation, and last follow-up, along with postoperative complications
Case
1 2 3 4 5 6 7
Gender, age (years)
ADL score of JOA Preop
3 months p/o
Last followup
Collapse of graft
Migration of pedicle screws
Instability of adj. level
Collapse of adj. vertebra
F, 56 F, 65 M, 61 F, 67 F, 74 M, 71 F, 59
1 2 1 5 3 2 2
6 7 6 8 7 5 7
2 7 1 8 – 3 6
– – + – – – –
– – + – – + –
– + – – – + +
+ – + – + + –
Case 4 A woman aged 67 years old compained of low-back pain and cauda equina intermittent claudication. We performed double PLIF on L3/4 and L4/5 and used aa Branigan I/F cage for each level. At 1 year after the surgery, we could confirm fusion of the graftsand no loosening of the pedicle screws (Fig. 4). She recovered her gait ability owing to improvement of the low-back pain and cauda equina symptoms.
Results Stable fixation of the lumbar spine was obtained after surgery. We confirmed fusion of the grafts in 85% of patients. All of the patients showed a marked improvement in their activities of daily living before they were discharged due to the relief of low-back pain and cauda equina symptoms. However, during follow-up, low-back pain increased and gait disturbance recurred in 3 patients who all suffered from the mutilated type of RA. Evaluation by ADL score in the JOA score system averaged 2.3, 6.6, and 4.5 at preoperation, 3 months after surgery, and last follow-up, respectively (Table 2). The deterioration on last follow-up was attributed to complications. In particular, adjacent vertebral collapse occured in 4 out of 5 mutilated-type RA patients (Table 3).
Discussion In 1952, Baggenstoss et al. [1] first reported on rheumatoid spondylitis of the lumbar spine and demonstrated pathologically typical rheumatoid nodules between the T12 and L3 vertebral bodies. In 1961, Seaman [12] and Lorber [9] reported a case of vertebral mutilation and osteolytic lesions. In 1964, Lawrence et al. [7] reported that 50 RA patients, when compared with 50 controls, showed a sigTable 3 Complication rate Collapse of adj. vertebra Instability of adj. level Migration of pedicle screw Collapse of grafted bone Decubitus and infection
57% (4/7) 43% (3/7) 29% (2/7) 14% (1/7) 14% (1/7)
Complication
nificantly higher incidence and severity of stenosis of the vertebral body without lipping and a significantly higher incidence of apophyseal joint mutilation. In 1978, Shichikawa et al. [13] reported that biopsies performed on 4 RA patients revealed 2 cases of rheumatoid spondylitis of the lumbar spine. Rheumatoid spondylitis of the lumbar spine requires distinction from findings associated with degenerative intervertebral disk and osteoporosis and from pyogenic spondylitis. In our cases, CT revealed the articular processes and vertebral erosion characteristic of RA. We also observed that lesions appeared less bright on T1-enhanced and T2-enhanced MR images from the vertebral endplate to the vertebral body. These observations differ from the findings of pyogenic spondylitis. During the 1980s, opinion varied widely on the best surgical treatments for rheumatoid spondylitis of the lumbar spine, but only a few reports were published. In 1984, Tsuji [15] compared RA with ankylosing spondylitis (AS): AS is repaired more quickly, and RA is often progressive. RA lesions include osteoporosis, collapse, cartilage erosion, and subluxation accompanied by intervertebral stenosis. Based on his cases, he concluded that fusion is suitable in the presence of low-back pain with neurological complications and ADL disorders. There were some reports on surgical cases: in 1989, Tajima et al. [16] reported on procedures performed in RA cases, including a laminectomy and anterior fixation, one laminectomy and posterior fusion, and one posterior fixation. They also indicated that, based on their experience, anterior fixation alone was not effective in cases of severe lesions of the vertebral body, because of progressive collapse of the vertebral body after anterior fusion surgery; in 1992, Nojiri et al. [17] reported that fusion was performed in 3 RA patients, one intervertebral PLIF and one ISOLA between multiple vertebral spines. In 1995, Nishioka et al. [18] reported on the pathology and prognosis of rheumatoid spondylitis of the lumbar spine, which was classified into two groups: the vertebral body mutilation group and the decayed apophyseal joint group. Collapse of the vertebral body occurred most often from the lower thoracic spine to the upper lumbar spine, while lesions in the apophyseal joint and intervertebral disk occurred more frequently in the lower lumbar spine. All of our patients showed lesions in the lower lumbar spine.
78 Fig. 5 Microscopic examination showed destruction of the apophyseal joint and vertebral endplate due to pannus, and infiltration of lymphocytes and plasmocytes Fig. 6 In the non-mutilated type, support in the early postoperative period was maintained by short fixation with a pedicle screw system. In the mutilated type, PLIF with a long fixation system or another stabilizing system and augmentation of the vertebral bodies might be needed
We selected PLIF as the common procedure for these 7 patients with rheumatoid spondylitis of the lumbar spine while we were preparing our surgical treatment plan. The common pathology in these 7 cases was the lesion caused by intervertebral instability due to mutilation of the apophyseal joint and intervertebral disk. Arthritis developed due to RA lesions, peridiscal enthesitis was added to it, and articular mutilation made irreversible progress. Microscopic examination showed destruction of the apophyseal joint and vertebral endplate due to pannus, and infiltration of lymphocytes and plasmocytes (Fig. 5). The patients’ pathology indicated the need for restoring the support of the anterior elements and for decompressing the entire circumference of the posterior neurological elements. PLIF alone can satisfy these requirements. Preservation of the vertebral endplate, strength of the grafted bone, and formation of a grafted bone matrix are the basic conditions leading to a successful PLIF for the non-mutilated type. In the non-mutilated type, fusion of the grafted bone and matrix was observed, and support in the early postoperative period was maintained by short fixation with a pedicle screw system. In the mutilated type, however, the endplates are mutilated, and osteoporosis has already progressed, and collapse of the graft, migration of the pedicle screws, instability of the adjacent level, and collapse of the adjacent vertebra were noted. There-
fore, PLIF with a long fixation system or another stabilizing system and augmentation of the vertebral bodies might be needed (Fig. 6). PLIF with spinal instruments is the treatment of choice for rheumatoid spondylitis of the lumbar spine, but in the mutilating type of RA with severe osteoporosis, PLIF in combination with a long fixation system and/or augmentation of the vertebral bodies might be needed.
References 1. Baggenstoss AH, Bickel WH, Ward LE (1952) Rheumatoid granulomatous nodules as destructive lesion of vertebrae. J Bone Joint Surg Am 34: 601–609 2. Biasi D, Caramashi P, Carletto A, et al (1995) A case of rheumatoid arthritis with lumbar spine involvement. Rheumatol Int 15: 125–126 3. Friedman H (1970) Intraspinal nodule causing nerve root compression. J Neurosurg 32: 689–691 4. Hauge T, Magnes B, Skullerud K (1980) Rheumatoid arthritis of the lumbar spine leading to anterior vertebral subluxation and compression of the cauda equina. Scand J Rheumatol 9: 241–244 5. Helliwell PS, Zebouni LNP, Porter G, Wright V (1993) A clinical and radiological study of back pain in rheumatoid arthritis. Br J Rheumatol 32: 216–221 6. Heywood AWB, Meyers OL (1986) Rheumatoid arthritis of the thoracic and lumber spine. J Bone Joint Surg Br 68: 362–368
79 7. Lawrence JS, Sharp J, Ball J, Bier F (1964) Rheumatoid arthritis of the lumbar spine. Ann Rheum Dis 23: 205–217 8. Linquist PR, McDonnell DE (1970) Rheumatoid cyst causing extradural compression: a case report. J Bone Joint Surg Am52: 1235–1240 9. Lorber A (1961) Osteolytic vertebral lesions as a manifestation of rheumatoid arthritis and related disorders. Arthritis Rheum 4: 514–532 10. Matsumine A, Shichikawa K, Yamashita K, et al (1998) Rheumatoid arthritis causing paraplegia. J Bone Joint Surg Am 34: 1410–1411 11. Nakase T,Fujiwara K, Kohno J, et al (1998) Pathological fracture of a lumbar vertebra caused by rheumatoid arthritis: a case report. Int Orthop 22: 397–399 12. Seaman WB (1961) Destructive lesion of the vertebral bodies in rheumatoid disease. Am J Roentogenol 86: 241–250
13. Shichikawa K, Matsui K, Oze K, et al (1978) Rheumatoid spondylitis. Int Orthop 2: 53–60 14. Sims-Williams H, Jayson MIV, Baddeley H (1977) Rheumatoid involvement of the lumbar spine. Ann Rheum Dis 36: 524–531 15. Tsuji A, et al (1984) Rheumatoid involvement of the lumbar spine. J Joint Surg 11: 65–70 16. Tajima N, et al (1989) Spinal instrument for the lumbar spine in rheumatoid arthritis. Jpn J Rheum Joint Surg 8: 205–212 17. Nojiri H, Matsui N, et al (1992) Surgical treatment for lumbar rheumatoid spondylitis. J Jpn Spinal Res Soc 3: 46 18. Nishioka J, et al (1995) Rheumatoid arthritis of the thoracic and lumbar spine. J Orthop Traum Surg 38: 209–218