S1 Failure in an Elderly Female
The patient is a 72-year-old female who was treated for spinal stenosis and scoliosis 3-years ago. She underwent a T10-S1 fusion at another institution. The patient is 5'2" tall and weighs 252 pounds. She did well for 2-months after surgery and then, progressively, back pain increased and posture changed.
The patient has no neurological deficits upon examination.
The patient's lateral supine x-ray (Figure 1A) and AP supine x-ray (Figure 1B) are below. (Both before a T10-S1 fusion at another institution)
Figure 1A. Lateral supine x-ray
Figure 1B. AP supine x-ray
The patient's postoperative (after T10-S1 fusion at another institution) lateral supine x-ray (Figure 2A) and AP supine x-ray (Figure 2B) are below.
Figure 2A. Lateral supine x-ray
Figure 2B. AP supine x-ray
Below (Figure 3) is the patient's lateral x-ray at 6-months postoperative to the T10-S1 fusion.
Figure 3. 22.8 cm PSB
Figure 4. Axial CT scan shows failure of S1 screws
Figure 5. Sagittal CT shows reconstruction
Failure of S1 fixation and likely multilevel pseudarthrosis. The patient has significant positive sagittal imbalance.
Suggest TreatmentIndicate how you would treat this patient by completing the following brief survey. Your response will be added to our survey results below.
At a different institution, the patient underwent a L4-L5 and L5-S1 ALIF and re-instrumentation with T10 to iliac fusion. (Figures 6A, 6B)
Figure 6A. AP x-ray L4-L5 and L5-S1 ALIF, T10 to iliac fusion
Figure 6B. Lateral x-ray L4-L5 and L5-S1 ALIF, T10 to iliac fusion
AP and lateral x-rays 1-year postoperative to the L4-L5 and L5-S1 ALIF and T10 to iliac fusion. (Figures 7A, 7B) A rod fracture is noted in Figure 7B.
Figure 7B. 17.9 cm PSB. Rod fracture is noted.
Figure 8. Axial CT myelogram at L5-S1 demonstrates pseudarthrosis
The patient did well for 9-months. Thereafter, back pain worsened and posture changed. There was an iliac screw prominence.
The patient then underwent an L5-S1 TLIF, T3 to iliac with revision iliac screws, and a L2 pedicle subtraction osteotomy. (Figures 9A, 9B).
At 1-year, the patient had marked improvement in posture, mild reduction of low back pain, and an Oswestry Disability Index score of 34.
This long adult spinal reconstruction highlights many important points that create a very difficult and complicated surgical course for some of our patients.
Figures 1A and 1B shows the pre-operative films of a 72-year-old female with a mild degenerative scoliosis and multilevel degenerative sagittal plane alignment. Of importance is that these short films do not allow visualization of the entire spinal column from the cervical spine to the pelvis. Also, these x-rays are taken supine and not upright, so there is no information on what the overall sagittal balance is pre-operative. Surgeons contemplating long spinal reconstructions, from the thoracolumbar junction to L4, L5 or the sacrum, need to obtain pre-operative upright long cassette AP and lateral radiographs. Such radiographs enable the surgeon to fully assess the entire spinal column, pelvis, and alignment to optimize postoperative alignment and balance.
Figures 2A and 2B shows again short cassette AP and lateral supine x-rays with a pedicle screw/rod construct for the thoracolumbar junction to S1. Of note, is the S1 screws are placed in a straight-ahead position on the AP x-ray and are short of being bicortical on the lateral x-ray. It is imperative to obtain strong bicortical purchase of S1 screws at the end of long constructs.
Also, acquiring long cassette x-rays to assess overall alignment and balance should be accomplished before the patient leaves the hospital. The ability of S1 screws alone to remain secure in this type of construct, without being supported by S2 or iliac screws posteriorly, or any interbody support of L5-S1 anteriorly, is very challenging.
Predictably, the S1 screws will migrate/loosen in a cantilever flexion type of moment arm, especially when such a tall L5-S1 disc is encountered. The unfortunate sequelae is noted clinically in Figure 3, which shows a mid-length lateral x-ray with the S1 screws nearly completely pulled out of the sacrum. The CT scans (Figures 4 and 5) confirm this. Now, a situation more difficult than the original problem presents in that a revision reconstruction is needed to realign and re-instrument the posterior spine.
Figures 6A and 6B finally shows long cassette AP and lateral x-rays following a revision ASF/PSF with interbody allograft spacers placed at L4-5 and L5-S1 and a revision PSF to the sacrum, including iliac screws. The lumbosacral region is much more secure, but the continued anterior global sagittal imbalance is still worrisome. This produces tremendous cantilever forces on the distal L5-S1 construct during standing and gait (often 6 to 7 times the patient's body weight), with the posterior fusion bone in continuous tension.
Also, no mention is made of the type of bone graft utilized in the most recent reconstruction. In a manuscript from our institution several years ago, our pseudarthrosis rate, at L5-S1 in long constructs to the sacrum/ilium, was greater than 25%. (1) Thus, Figures 7A and 7B, which shows the 1-year postoperative x-rays, demonstrates broken L5-S1 rods and suspected lumbosacral pseudarthrosis with recurrent sagittal imbalance. This is a combination of failed lumbosacral instrumentation/pseudarthrosis and proximal junctional kyphosis above the prior thoracic fusion levels. This was appropriately treated with a revision PSF with implant removal, L2 PSO and PSF from the upper thoracic spine to the sacrum/ilium.
Figures 9A and 9B, 1-year postoperative x-rays, show improved sagittal balance, but with a persistent and approximate 8-10 cm sagittal vertical axis. The outcome scores are improved, but a fair amount of disability remains.
Also, it is imperative to follow these patients for at least 5-, if not 10-years after surgery. Many adult pseudarthrosis patients will be demonstrated after 2-years, and many after 5-years follow up, as shown in the paper by Kim et al JBJS, 2006 (1).
So, although the alignment and construct are quite satisfactory at the latest follow up, this now mid-70-year-old lady, who has had 2 prior lumbosacral reconstructions for pseudarthrosis, and implant failure is not without risk of further problems, will require longer follow up. This case demonstrates many of the challenges of long spinal reconstructions in an elderly age group. In addition, when ending a long construct from the thoracic spine to sacrum, the use of only 2 S1 screws, without any backup posterior fixation, such as iliac screws, or any interbody support placed via a TLIF or ALIF route, is also quite prone to clinical failure.
We are presenting a series of 33 patients revised at our institution for S1 screw failures/L5-S1 pseudarthrosis in long adult spinal reconstructions at the 2008 Scoliosis Research Society meeting in Salt Lake City (2). In this presentation, 17 out of 19 patients with only S1 screws and no other distal fixation had failure of the screws at L5 and/or S1. Even if the L5-S1 interspace had been structurally grafted, 4 of 6 bone grafts collapsed and 2 of 12 cages subsided. In those with bilateral S1 and /or bilateral S2 or Iliac screws, only L5-S1 rod failure was noted demonstrating lumbosacral pseudarthrosis. All patients revised had bilateral bicortical S1 screws and at least 1 iliac screw and usually 2 placed. Fifteen out of 21 of these revisions ultimately healed, but 6 needed additional surgery to become solid. Hence, the difficulty in this patient population.
1. Kim YH, Bridwell KH, Lenke LG, Cho SK, Edwards II CC, Rinella AS: Pseudarthrosis in Adult Spinal Deformity Following Multisegmental Instrumentation and Arthrodesis. JBJS 2006: 88-A(4): 721-8.
2. Harimaya K, Mishirao T, Lenke LG, Bridwell KH, Koester L, Sides B: Etiology and Revision Surgical Strategies in Failed Lumbosacral Fixation of Adult Spinal Deformity Constructs. Scoliosis Research Society Annual meeting, Paper # 23, Salt Lake City, UT, September 2008.