Stabilized C-reactive protein-albumin-lymphocyte (CALLY) index predicts adjacent fractures after kyphoplasty.

objectiveTo determine whether a stabilized C-reactive protein–albumin–lymphocyte (CALLY) index measured 4–6 weeks after percutaneous kyphoplasty (PKP) independently predicts adjacent vertebral fracture (AVF), and to compare its discrimination with other immuno-nutritional indices.

methodsConsecutive patients undergoing single-level PKP for osteoporotic vertebral compression fracture at a tertiary center between January 2020 and December 2024 were analyzed. Stabilized laboratory tests were obtained 4–6 weeks postoperatively. CALLY was calculated as albumin (g/L) × lymphocyte count (×10⁹/L) divided by C-reactive protein (mg/L). The primary outcome was incident AVF during follow-up. Multivariable logistic regression adjusted for age, sex, body mass index, prior fragility fracture, lumbar spine T-score, intradiscal cement leakage, and kyphosis correction. Discrimination was assessed by receiver operating characteristic curves with DeLong comparisons. Cut-offs were derived using Youden’s J. Pre-specified subgroup analyses evaluated effect modification.

resultsOf 512 patients, 102 (19.9%) developed AVF. The AVF group had a lower stabilized CALLY than the non-AVF group (median 9.9, IQR 6.8–15.9 vs. 12.9, IQR 9.0–20.4; p < 0.001). In the adjusted model, each 5-unit decrease in CALLY was associated with higher odds of AVF (adjusted odds ratio 1.29; 95% CI 1.13–1.49; p < 0.001). Lower lumbar spine T-score, intradiscal cement leakage, and greater kyphosis correction were also independent predictors. As stand-alone biomarkers, discrimination ranked CALLY area under the curve (AUC) 0.691, neutrophil-to-lymphocyte ratio 0.592, prognostic nutritional index 0.587, and platelet-to-lymphocyte ratio 0.563. A clinical–procedural base model achieved AUC 0.632. Adding CALLY increased AUC to 0.714, with a significant DeLong difference versus the base model (ΔAUC 0.082; 95% CI 0.004–0.160; p = 0.0389). DeLong tests also favored CALLY over other single markers. A CALLY cut-off of 13.0 yielded sensitivity 0.69 and specificity 0.49. The association was consistent across most subgroups and was stronger with intradiscal cement leakage (interaction p = 0.03).

conclusionsA lower stabilized CALLY index measured 4–6 weeks after PKP identifies patients at increased risk of adjacent fracture and provides incremental discrimination beyond standard clinical and procedural predictors. Incorporating CALLY into early postoperative assessment may support targeted secondary prevention.