Adult spinal deformity with low pelvic incidence: a challenging condition demanding a tailored strategy

Existing surgical alignment goals derived from populations with a high pelvic incidence (PI) are not applicable for patients with adult spinal deformity (ASD) and a low PI, who account for a high proportion of Asian populations. The surgical treatment for patients with a low PI is challenging because of their limited pelvic compensation capacity and because there are no criteria to guide corrective spinal deformity surgery in this population.

To develop and validate a tailored sagittal correction strategy for patients with ASD and a low PI.

Cross-sectional normative analysis and retrospective cohort study.

Stage I included 852 asymptomatic Chinese adults (age 50-79 years). Stage II included 103 patients with ASD and a PI of ≤41° who underwent posterior long-segment fusion and follow-up evaluation for ≥2 years, stratified by kyphotic apex into a TL group (kyphotic apex at L1 or above; n=59) and an L group (kyphotic apex at L2 or below; n=44).

Mechanical complications (MCs) and health-related quality of life (HRQOL), as measured by the Oswestry Disability Index and a visual analog scale for back and leg pain.

Normative spinopelvic parameters were used to define a low-PI subgroup (PI ≤ 41°) and to derive the sufficient sacral slope-lumbar lordosis matched correction (SSS-LLMC) strategy. The 25th percentile of sacral slope (SS) in asymptomatic adults with a low PI (41°) was adopted as the minimal SS target (SS ≥ 21°). Patients with ASD and a low PI were divided into the sufficient sacral slope (SSS, postoperative SS ≥ 21°) group and the insufficient sacral slope (ISS, postoperative SS < 21°) group. According to the linear sacral slope-lumbar lordosis (SS-LL) relationship (LL=1.122 × SS + 10.84) established in asymptomatic adults with a low PI, patients in SSS group was further stratified into lumbar lordosis matched correction (LLMC), lumbar lordosis undercorrection (LLUC), and lumbar lordosis overcorrection (LLOC). MCs and HRQOL over a minimum 2-year follow-up period were compared across these subgroups within the TL and L groups, and multivariate logistic regression identified independent predictors of MCs in the overall and apex-stratified cohorts. In addition, this strategy was compared with conventional alignment goals, such as the Scoliosis Research Society-Schwab modification of the pelvic incidence to lumbar lordosis (PI-LL) mismatch and the global alignment and proportion (GAP) score, to evaluate the ability of these approaches to reduce MCs.

A low PI (≤ 41°) accounted for 32.2% (274/852) of the asymptomatic cohort. Among 103 patients with ASD and a low PI, 36 (35.0%) developed MCs. MCs occurred in 56.8% (20/34) of patients with an ISS versus 23.2% (16/69) of patients with a SSS (p<.001). Within the SSS group, MC rates were 47.4% (9/19) in LLUC, 8.6% (3/35) in LLMC, and 26.7% (4/15) in LLOC (p=.004). In the TL group, SSS-LLMC had the lowest MC rate (12.5%; 3/24), whereas in the L group no MCs occurred in SSS-LLMC (0/10) and the MC rate in SSS-LLOC (14.3%; 1/7) was lower than in SSS-LLUC (57.1%; 4/7). Preoperative SVA was an independent risk factor for MCs (odds ratio [OR]=1.224; 95% confidence interval [CI], 1.088-1.377; p<.001), and SSS-LLMC was independently protective (OR, 0.116; 95% CI, 0.031-0.435; p=.001). In the L group, SSS alone was independently protective (OR, 0.187; 95% CI, 0.047-0.753; p=.018). At the final follow-up evaluation, the ODI was similar between the ISS and the SSS groups. In the TL group, the SSS group showed a lower mean VAS for the back compared to ISS group (2.7±1.0 vs 3.4±1.0; p=.028), while all other between-group comparisons yielded no statistically significant results.

Patients with ASD and a low PI who require posterior long-segment fusion treatment represent a distinct anatomic subtype requiring individualized correction targets. The SSS-LLMC strategy prioritizes restoring SS as a foundation before precisely matching LL to the restored pelvic orientation. In addition, the target for LL reconstruction should take into account the different locations of the kyphotic apex.