OTSR - 2026-03-19 - Journal Article
An archetypal analysis of lumbopelvic profiles could help predict adverse spinopelvic mobility after total hip arthroplasty.
Aubert T, Hallé A, Aubert O
Topics
Key Takeaway
Lumbopelvic archetypes associated with sagittal imbalance and lumbar stiffness (A4–A7) had rates of adverse postoperative spinopelvic mobility (ΔSPT ≥20°) of 60–87.5%, with adjusted ORs of 3.2–7.5 versus lower-risk archetypes.
Summary Depth
Choose how much analysis to show on this article page.
Summary
This study applied a previously trained 7-archetype unsupervised machine learning model to 108 anterior-approach THA patients to determine whether preoperative lumbopelvic archetype predicts adverse postoperative spinopelvic mobility (ΔSPT ≥20°). Preoperative parameters (SPT, LL, PI, lumbar flexion, PI-LL mismatch) were used to classify patients, with standing and flexed-seated lateral radiographs obtained pre- and 3 months postoperatively. Archetypes A4–A7, characterized by sagittal imbalance and lumbar stiffness, carried 60–87.5% rates of adverse mobility and adjusted ORs of 3.2–7.5 on logistic regression (p<0.05).
Key Limitation
Three-month follow-up is insufficient to determine whether adverse spinopelvic mobility patterns persist, resolve, or translate into clinically meaningful endpoints such as dislocation or impingement.
Original Abstract
BACKGROUND
The hip‒spine relationship (HSR) plays a crucial role in the outcomes of total hip arthroplasty (THA) as it influences spinopelvic mobility and the risk of prosthetic impingement. Traditional preoperative assessments, which rely on static spinopelvic parameters, often fail to predict postoperative changes. To overcome this limitation, we developed an unsupervised machine learning model on the basis of an archetype analysis that revealed seven distinct lumbopelvic profiles, independent of modifiable postoperative factors. The aim of this study was to apply the trained model to a new patient cohort and analyse the postoperative spinopelvic mobility patterns across archetypes.
HYPOTHESIS
Our hypothesis was that the use of an archetypal approach would allow us to anticipate patients with changes in lumbopelvic mobility and those at risk of adverse spinopelvic mobility.
METHODS
The clinical data of 108 patients who underwent THA via the anterior approach by a single surgeon were analysed retrospectively. The preoperative spinopelvic parameters-age, standing pelvic tilt (SPT), lumbar lordosis (LL), pelvic incidence (PI), lumbar flexion (LF), and PI‒LL mismatch-were input into the previously trained archetype analysis model. Patients were classified into one of seven archetypes on the basis of their highest archetype score (summing to 1.0 per patient). Spinopelvic mobility (ΔSPT) and the pelvic femoral angle (ΔPFA) were assessed from standing and flexed-seated lateral radiographs preoperatively and three months postoperatively. Adverse spinopelvic mobility was defined as a ΔSPT ≥ 20°. Associations between archetypes and postoperative mobility were evaluated via logistic regression.
RESULTS
The mean postoperative ΔSPT increased by 9.53° (-34.4° to 50.3°), with the rate of a postoperative ΔSPT ≥ 20° increasing significantly in archetypes associated with sagittal imbalance and lumbar stiffness: A4 (63.33%), A5 (60%), A6 (62.5%), and A7 (87.5%) (p = 0.01). Logistic regression confirmed that these archetypes were associated with the highest risk of postoperative adverse mobility (adjusted ORs ranging from 3.2 to 7.5; p < 0.05).
CONCLUSION
This AI-driven archetypal classification demonstrates an association with adverse spinopelvic mobility, suggesting potential value for risk stratification, identifying high-risk profiles that may benefit from personalized surgical strategies, including implant orientation modifications.
LEVEL OF EVIDENCE
IV; retrospective study.