Early fidelity and long-term drift: do we achieve preoperative alignment goals in adult spinal deformity?

To determine how often adult spinal deformity (ASD) surgery meets preoperative sagittal alignment goals and whether these corrections are durable over time, including the impact of planning frameworks and enabling technologies.

We performed a PRISMA-guided search of MEDLINE, EMBASE, Web of Science, Cochrane Library, and Scopus through June 2025. Eligible studies enrolled adults undergoing ASD surgery where preoperative alignment targets were defined by numeric thresholds (e.g., SVA, PI-LL, PT), algorithmic frameworks (e.g., Roussouly, GAP), or technology-assisted modalities (e.g., virtual surgical planning/patient-specific rods/robotics/custom implants). Outcomes included plan-to-achieved alignment fidelity, longitudinal maintenance of targets, mechanical complications, and PROMs. Risk of bias was assessed with ROBINS-I. When data were sufficiently homogeneous, random-effects meta-analyses were performed for commonly reported targets (SVA < 5 cm, PI-LL ≤ 10°, PT ≤ 20°).

Fifteen unique studies (Fig. 1) from 2006-2025 encompassing ~ 1,980 patients reported outcomes for achieving preoperative alignment goals in ASD. Cohort sizes ranged from 15 to 608; among cohorts reporting demographics, mean age was 65 years (range means 58-72) with 68% female (57-80%). Mean follow-up was 20 months (immediate postoperative to 5 years). Across studies, most patients achieved early postoperative targets for SVA, PI-LL and/or PT, though durability attenuated on longer follow-up in several cohorts. Algorithm-driven planning-particularly restoration to a Roussouly-consistent profile-was associated with lower mechanical complication rates, while technology-assisted approaches (virtual planning/patient-specific rods/custom implants/robotics) generally improved plan-to-achieved fidelity; translation to superior PROMs was inconsistent. Overall study quality varied: several small, single-center series were at higher risk of bias, whereas multicenter prospective cohorts showed lower risk but remained heterogeneous in definitions, targets, and follow-up windows. Fig. 1 Prisma chart [17] CONCLUSIONS: Adult spinal deformity surgery reliably achieves planned sagittal alignment early postoperatively, but long-term durability is variable. Pelvic tilt is the most difficult parameter to correct and maintain, with loss of correction by one to two years. Alignment strategies emphasizing restoration of physiologic sagittal profile, particularly Roussouly-based frameworks, show superior long-term maintenance and fewer mechanical failures than isolated numeric targets. Although patient-reported outcomes improve consistently, enabling technologies enhance alignment fidelity without consistent incremental PROM benefit. Future studies should prioritize profile-driven, patient-specific planning and assess durability and patient-centered outcomes beyond two years in prospective multicenter cohorts.