The Natural History of Leg Pain after Laminectomy vs. Transforaminal Interbody Fusion with Bone Morphogenic Protein
Lumbar spinal stenosis can produce pain with varying distance down the leg: to the buttock, posterior thigh, or past the knee (radicular). Surgical decompression usually improves these symptoms, though in some patients, similar pain can remain or occur new. Both the TLIF procedure and off label use of rhBMP-2 (in TLIF) have also been complicated by new radiculopathy, whether from the decompression, retraction of the nerve root, trauma to the dorsal root ganglion, or irritation of neural tissue caused by the BMP. The natural history of residual leg pain or new leg pain after laminectomy or from TLIF with BMP is unknown.
This is the first study to analyze occurrence, distribution, and natural history of all types of leg pain after laminectomy compared to TLIF using interbody BMP.
Retrospective review of prospectively collected data
495 consecutive adults age 61 years (19-91 years) underwent open laminectomy or laminectomy and TLIF+BMP
Leg pain presence and distribution (buttock only, thigh, past knee) was analyzed pre-op and post-op 6 weeks, 3 months, 6 months, 1 year and 2 years. Pain scores (VAS) were compared with the sign test. Oswestry (ODI) differences tested for normalcy using Anderson Darling, compared using paired t-tests. Outcomes evaluated with chi square.
All patients had lumbar stenosis on MRI. Pain distribution was characterized from history and pain diagrams. Excluded: simple disc herniations, anterior surgery. Fusion indications: degenerative, spondylolisthesis, or deformity. Laminectomy only group: 74 patients (59 primary, 15 revision) lami at 3.0 levels (1-5). The 421 patients (255 primary, 166 revision) in the TLIF+BMP group required laminectomy 2.6 levels (1-6), TLIF 1.7 levels (1-4), and posterior instrumented fusion 4.3 levels (2-17).
Follow-up averaged 65 months (24-117 months). Lami group: VAS improved from 5.7 pre-op to 2.8 (p=0.026) at 6 weeks, 2.6 at 3 months (p=0.042) and remained stable. ODI improved from 49 pre-op to 24 (p<0.001) at 6 weeks, 28 at 3 months (p=0.049) and remained stable. TLIF group: VAS improved from 6.2 pre-op to 3.9 (6 weeks), 3.5 (3 months), 3.1 (6 months), 2.9 (1 year), 3.1(2 year); ODI improved from 49 pre-op to 32 (6 weeks), 34 (3 months), 28 (6 months), 26 (1 year), 28 (2 year) (p<0.001). Leg pain past the knee in 416 TLIF group was still present in 19% at 6 weeks, 13% at 3 months (8% for primary TLIF), and remained stable. Lami group leg pain past the knee was present in 10% at 6 weeks, 11% at 3months, and remained stable (7% for primary lami). Both groups achieved similar long-term VAS (p=0.152). Both groups had similar levels of revision surgery, with slightly more in the TLIF group (38% vs. 32%).Transient increase in buttock or thigh pain at 6 weeks was seen in revision laminectomy (26%), revision TLIF (12%), and primary TLIF (1%), each returning to baseline by 3 months.
Comparing laminectomy vs. TLIF+BMP in patients with leg pain, TLIF had a small transient increase in leg pain at 6 weeks, with both groups similar at 3 months. Revision surgery was associated with more residual leg pain short and long-term, without difference between groups. Post-op buttock/thigh pain showed transient increase in both groups. Lami group clinical outcomes were stable after 6-12 weeks, with TLIF group stabilizing after 6 months.