Minimum Two-Year Analysis of Sacropelvic Fixation and L5/S1 Fusion Utilizing S1 and Iliac Screws

Kuklo TR,
Bridwell KH,
Lewis SJ,
Baldus C,
Blanke K,
Lenke LG
St. Louis, MO, USA

INTRODUCTION:
S1 screws often fail with lumbosacral fusions and L5/S1 pseudos are common in deformity patients (pts). The purpose of this study was to assess the clinical and radiographic results of iliac/S1screw constructs used for long fusions to the sacrum and high–grade spondylolisthesis at a single institution.

MATERIALS AND METHODS:
Between 1993 and 1998, 81 pts (38 revision, 43 primary) with minimum 2–year follow–up (ave. 4.2 yrs, range 2.0–7.1) underwent sacropelvic fixation and L5/S1fusion using S1 and iliac screws (154 screws). 3 generations of Liberty iliac screws (range 6.0–7.5 mm diameter/60–70 mm length) were used with bicortical S1 pedicle screws (range 5.5–7.0 mm diameter/35–45 mm length). 49 of 81 constructs (61%) included an anterior load sharing/fixation device (titanium mesh cage – 29, transacral allograft fibula – 12, fresh frozen allograft femoral ring – 4, BAK cage – 2, autograft fibula – 1, and tricortical iliac crest graft – 1). Group I included isthmic spondylolisthesis (n=42) fused L4 or L5 to sacrum; Grade II (7), Grade III (19), Grade IV (10), Grade V (6). Group 2 included long fusions (>3 levels) to the sacrum (n=39) – adult lumbar scoliosis (13), fixed sagittal imbalance syndrome (12), multiple level degenerative disc disease (8), congenital scoliosis (3), neuromuscular scoliosis (2) and aneurysmal bone cyst resection (1). In Group 2, 15 pts (Group 2A) were fused from L1, L2 or L3 to the sacrum (3–5 level fusion, ave. 3.3. levels) and 24 pts (Group 2B) were fused from the thoracic spine to the sacrum (6–17 level fusion, ave. 11.5 levels). 38 pts (47%) presented after previous spine surgery: 22 (1 surgery) and 16 (2 or more surgeries). 12 pts presented with a pseudarthrosis at the lumbosacral junction. 6 of those 12 pseudos had both anterior and posterior surgery previously performed on presentation to us. Radiographs were analyzed for fusion at the lumbosacral junction slip angle, and sacral inclination if applicable, anterior loadshadng devices/constructs, sagittal balance, progressive iliosacral degenerative changes, screw size and position, and screw removal and complications. An 8–item patient questionnaire addressing pain, prominence and function was also completed.

RESULTS:
36 of these 38 revision pts had previous iliac bone grafting, yet iliac screws were successfully placed in 34 of the 36 previously–harvested graft sites. On follow–up, 4 of the 81 pts (4.9%) had pseudarthrosis at L5/S1 after reconstruction by us (2 Group 1 pts and 2 Group 2B pts). 2 of those pts had anterior loadshadng/fixation at L5/S1 and 2 did not. Solid fusion was obtained in 10 of the 12 pts presenting to us with L5/Sl pseudarthrosis. 78 of the 81 pts had additional iliac crest bone graft harvesting (18 bilateral). None were noted to have a loss of screw fixation or iliac crest fracture after harvesting. 2 pts developed deep wound infections: 1 lost sacropelvic fixation (the only 1 in the series – a revision IIB pt – only 1 unilateral iliac screw), the other fused successfully. A small halo (<2 mm) was commonly seen around the iliac screws (47%) by 2 yr follow–up. However, this did not lead to screw failure or non–union. There was no premature sclerosis or degenerative changes noted in the sacroiliac joints. Based on the pt questionnaire, 47% of the pts thought the screws were prominent, but only 14% experienced some discomfort over the iliac screws secondary to prominence. 3% complained of buttock pain when setting >1 hour, 3% buttock pain when climbing stairs based on the questionnaire. 17 pts (21%), 15 of which were adolescents with solid fusion, eventually underwent elective iliac screw removal in a subsequent procedure.

CONCLUSIONS:
Bilateral iliac screws coupled with bilateral S1 screws appear to provide excellent distal fixation for lumbosacral fusions with a high fusion rate (95.1%). L5/S1 fusion rate was 83% in those with previous pseudos at L5/S1. There was overall a low complication rate. Previous iliac crest harvesting does not prevent ipsilateral screw placement (34 out of 36 pts) or additional iliac harvesting (78 out of 81pts).