18 Surgical Treatment of Lumbar and Thoracolumbar Curve Patterns (Lenke V)

Stephen G. George, Firoz Miyanji, and Harry L. Shufflebarger

Summary

Posterior spinal fusion utilizing segmental pedicle screws and multilevel Ponte osteotomy has become the standard in the surgical treatment of adolescent idiopathic thoracolumbar and lumbar scoliosis. This is superior to the anterior thoracolumbar procedure in every modality except the number of vertebrae fused. The surgical technique has evolved in the past two decades. Currently, convex rod placement first begins in the apical vertebra and progresses to the distal and then to the proximal ends, with derotation and compression at each level. Choosing fusion levels for the posterior procedure is relatively standard. Fusion levels are usually Cobb-to-Cobb measurements. The lower instrumented vertebra (LIV) relates to the L3–L4 disc. With L3 the end of the Cobb, the disc is wider on the right (in left curves) and L3 is clearly the LIV. With L4 at the end of the Cobb, the disc is wider on the left and L4 is the LIV. With the disc parallel, the decision is more problematic. The senior author (H.L.S.) usually chooses L3 in this situation. However, if L3 (the intended LIV) is less than two disc spaces distal to the apex of the curve, L4 should be chosen.

Key words

lumbar/thoracolumbar scoliosis – posterior spinal fusion – Lenke V curves – lumbar/thoracolumbar deformity correction – fusion level selection

18 Surgical Treatment of Lumbar and Thoracolumbar Curve Patterns (Lenke V)

18.1 Introduction

Harrington instrumentation was the first truly effective spinal instrumentation system, becoming the standard of care for all scoliosis surgery regardless of type. As time went on, the deleterious effects of posterior distraction in the lumbar spine became evident (the “lumbar flatback syndrome” ¹ ) requiring frequent revision and extension distally. Allan Dwyer and Klaus Zielke both introduced anterior instrumentation systems for the correction of Lenke V curves. ² ^, ³

Unfortunately, these proved to be uniformly kyphogenic in the lumbar spine due to the anterior convex compression. Halm–Zielke instrumentation and Kaneda instrumentation developed variations to attempt to avoid this effect. ⁴ ^, ⁵

Jürgen Harms introduced the concept of anterior load sharing utilizing a titanium mesh cage filled with graft as an interbody device with anterior instrumentation in Lenke 5 curves, uniformly maintaining lumbar lordosis and enhancing coronal and axial correction. ⁶ This concept and technique remains the foundation of anterior surgery for Lenke 5 deformities.

Current anterior surgery for Lenke 5 involves a thoracolumbar transthoracic retroperitoneal approach to the spine, usually from T10 to L3 or 4. This has been reported to have significantly longer surgical times and length of stay compared to posterior approaches. Kim et al ⁷ reported on the morbidity of this anterior approach in adult spinal deformity patients. This was associated, at 5-year follow-up, with a high rate of pain (32%), abdominal bulging (43%), and functional disturbance (24%). They counseled using caution when recommending this approach to spinal deformity patients. Although these undesirable results have not been reported in an adolescent population, it is likely that they similarly exist.

Understandably, the posterior approach was largely abandoned for Lenke 5 curves in the 1980s and 1990s. Due to poor results with anterior devices, the senior author (H.L.S.) utilized a wide posterior release in conjunction with Cotrel–Dubousset hooks to maintain or increase lumbar lordosis in the late 1980s. ⁸ With the advent of posterior segmental pedicle screw instrumentation and the Ponte osteotomy, ⁹ the changing publication practice of this technique in Lenke 5 curves served as the basis for the resurgence of posterior instrumentation in Lenke 5 curves. This was followed by a matched cohort comparison of the anterior dual rod with interbody spacer instrumentation (St. Louis) versus posterior segmental instrumentation with Ponte osteotomies (Miami). ¹⁰ Results from the posterior group were significantly better in multiple parameters: posterior correction, less correction loss, more lumbar lordosis production, shorter operative time, and shorter length of stay. Only levels instrumented were less for anterior, with no difference in distal instrumented level. Clearly, posterior surgery for Lenke 5 is superior to anterior surgery.

18.2 Level Selection and Surgical Technique

18.2.1 Upper and Lower Instrumented Vertebrae

In general, fusion levels in Lenke 5 curves are vertebrae inclusive in the Cobb angle measurement. The upper-end vertebra should be neutral, and is usually T10, occasionally T9.

The selection of the lower end vertebra is more controversial and problematic. Using the distal Cobb measurement, L3 or L4 will be the LIV. When L3 is the end of the Cobb, the L3–L4 disc space, in the usual left curve, is wider on the right than on the left. Here, the LIV is clearly L3. When L4 is the end of the Cobb, the L3–L4 disc space is wider on the left than on the right. Here, L4 is clearly the LIV. When the L3–L4 disc space is parallel, the decision is more difficult. It has been the senior author’s (H.L.S.) preference to stop at L3 in this situation. Miyanji et al, ¹¹ in a report of 161 cases of Lenke 5 curves, found in posterior approaches L3 was the LIV in 67% and L4 in 33%. In the anterior group, L4 was the LIV in only 4%. They also reported a significantly greater amount of disc angulation beneath the LIV in the group treated anteriorly at 2 years, with unknown long-term effects.

The relationship of the apex of the Lenke 5 curve to the intended LIV enters into the decision-making regarding level selection. It is preferable to have at least two disc spaces between the curve apex and the intended LIV. This is the senior author’s (H.L.S.) unsupported assumption.

18.2.2 When Does Lenke 5 Become 6?

The designation Lenke 6 denotes a structural coronal curve proximal to the lumbar curve on side bending or an increase in thoracic kyphosis. Several parameters can be utilized to define when a Lenke 5 becomes a Lenke 6. The clinical appearance of the patient is very important in this determination. In the absence of a significant thoracic rib prominence, acceptance of a higher degree thoracic curve may be entertained. Up to 45 degrees may be accepted in this situation. Asghar et al ¹² in an unpublished Harms Study Group report determined that 35 to 38 degrees was the upper limit of noninclusion of the proximal curve. It was also felt there was some degree of surgeon bias involved in this selection.

The amount of thoracic kyphosis for which inclusion of the thoracic curve is required is not well documented. It has been the senior author’s (H.L.S.) practice to include the proximal curve when thoracic kyphosis exceeds 60 degrees.

18.2.3 Ponte Osteotomy in Lenke 5 Curves

Ponte osteotomy has been an integral part of Lenke 5 surgery for the senior author (H.L.S.) for some 20 years. This in combination with segmental pedicle screws has resulted in outcomes equal to or better than any anterior surgery report. ⁹ ^, ¹⁰ Asymmetric osteotomy, wider on the convex side, is favored. The osteotomy can be done from pedicle to pedicle if necessary. The senior author’s (H.L.S.) preferred technique includes superior lumbar facet excision by osteotome or rongeur, excision of ligamentum flavum, excision of the anterior extension of the ligamentum flavum forming the anterior capsule of the facet joint with Kerrison rongeur, and excision of all or part of the inferior facet with Kerrison or Leksell rongeur. The osteotomy can be accomplished in less than 5 minutes. A hemostatic agent is placed in each osteotomy site to manage bleeding.

Osteotomy is usually accomplished from Cobb to Cobb, excepting the distal segment (Fig. 18‑1a–c) demonstrates the progression to completion of a Ponte osteotomy. Osteotomy is usually performed prior to screw placement.