Association of bone mineral density with reoperation rate following instrumented lumbar spinal fusion

Low bone mineral density (BMD) is believed to influence the outcome of instrumented spinal surgery and can lead to reoperation. Purpose of this observational and case-control study was to investigate the association of BMD with the risk of reoperation following instrumented lumbar spinal fusion (LSF). For the observational study, 81 patients were included who received LSF with and without augmentation. For the case-control study, 18 patients who had reoperation following LSF were matched to 26 patients who did not have reoperation (matched by sex, age +/- 5 years, fused levels and PMMA-augmentation). Opportunistic BMD screening was performed in perioperative CT scans using asynchronous calibration. Mean BMD was compared between patients with and without reoperation in augmented and non-augmented surgeries. In the observational study, prevalence of osteoporosis (BMD < 80 mg/cc) was 29% in non-augmented and 85% in augmented LSF. Seven of 48 patients with non-augmented (15%) and 4 of 33 patients with augmented LSF (12%) had reoperation. In non-augmented LSF, patients with reoperation had significantly lower BMD than patients without reoperation (P = 0.005). The best cut-off to predict reoperation after non-augmented LSF was BMD < 83.7 mg/cc. In the case-control study, patients with reoperation presented numerically lower BMD of 78.8 +/- 33.1 mg/cc than patients without reoperation with BMD of 89.4 +/- 39.7 mg/cc (P = 0.357). Despite much lower BMD surgeries with PMMA-augmentation showed no higher reoperation rate compared to non-augmented surgeries. Patients with reoperation following LSF showed slightly lower BMD compared to matched patients without reoperation, but the difference was not statistically significant. Opportunistic BMD screening can be performed in preoperative CT, thus informing about osteoporotic bone, a potential risk factor of surgery failure.

💡 Research Summary

This paper investigates the relationship between lumbar spine bone mineral density (BMD) and the risk of reoperation after instrumented lumbar spinal fusion (LSF). The authors employed two complementary study designs: an observational cohort of 81 patients who underwent LSF with either conventional (non‑augmented) pedicle screw‑rod fixation or polymethylmethacrylate (PMMA)‑augmented fixation, and a matched case‑control series comparing 18 patients who required reoperation with 26 patients who did not, matched on sex, age (±5 years), number of fused levels, and augmentation status.

In the observational cohort, 48 patients received non‑augmented constructs and 33 received PMMA‑augmented constructs. Osteoporosis (defined as BMD < 80 mg/cc) was present in 29 % of the non‑augmented group and 85 % of the augmented group, reflecting the older, more osteoporotic population selected for augmentation. Reoperation occurred in 7 of 48 (15 %) non‑augmented cases and 4 of 33 (12 %) augmented cases. Among the non‑augmented patients, those who required reoperation had a markedly lower mean BMD (73 ± 18 mg/cc) than those who did not (110 ± 38 mg/cc; P = 0.015). Receiver‑operating‑characteristic (ROC) analysis identified a BMD threshold of 83.7 mg/cc as optimal for predicting reoperation in the non‑augmented cohort, with an area under the curve of 0.798 (95 % CI 0.649–0.946; P = 0.013). This cut‑off closely aligns with the conventional diagnostic threshold for osteoporosis, suggesting that patients with BMD below this level are at heightened risk of construct failure when conventional fixation is used.

Conversely, the PMMA‑augmented group exhibited a substantially lower mean BMD (60 ± 24 mg/cc) yet did not experience a higher reoperation rate than the non‑augmented group. The augmented cohort also had longer fixation constructs (median 2 levels vs. 1 level in the non‑augmented cohort), a factor that may mitigate adjacent‑segment stress and contribute to the comparable reoperation rates. These findings support the biomechanical premise that PMMA augmentation improves screw purchase and overall construct stability in osteoporotic bone, thereby offsetting the increased failure risk associated with low BMD.

The case‑control analysis, designed to control for confounding variables, showed that reoperated patients had a numerically lower mean BMD (78.8 ± 33.1 mg/cc) than matched controls (89.4 ± 39.7 mg/cc), but the difference did not reach statistical significance (P = 0.357). The lack of significance likely reflects limited sample size, heterogeneity in the reasons for revision surgery, and the absence of a standardized, objective definition of “reoperation” in the dataset.

A notable methodological contribution of the study is the use of opportunistic quantitative CT (QCT) to derive volumetric BMD from routine pre‑ or immediate postoperative CT scans. The authors applied asynchronous calibration to convert Hounsfield units to BMD, incorporating correction factors for different tube voltages (120 kVp vs. 140 kVp) and for contrast phases (arterial and portal‑venous). This approach eliminates the need for dedicated dual‑energy X‑ray absorptiometry (DXA) scans, enabling bone health assessment without additional radiation or cost. However, the authors acknowledge that HU‑to‑BMD conversion is scanner‑specific, and multi‑center implementation would require standardized calibration phantoms or cross‑validation.

In the discussion, the authors contextualize their findings within the broader literature. Prior studies have linked low HU measurements on pre‑operative CT to higher rates of screw loosening, pseudarthrosis, and adjacent vertebral fractures. Biomechanical investigations have demonstrated that BMD below ~80 mg/cc compromises screw pull‑out strength and increases the likelihood of early construct failure. The present study adds clinical evidence that, in a real‑world surgical population, low BMD is indeed associated with a higher reoperation rate when conventional fixation is employed, and that PMMA augmentation can neutralize this risk despite markedly lower BMD values.

Limitations include the retrospective design, potential selection bias (patients with poorer bone quality may have been preferentially selected for augmentation), variable follow‑up durations, and the subjective nature of the reoperation endpoint (symptom‑driven versus radiographic criteria). The authors also note that loss to follow‑up could differentially affect the observed rates, although they argue that this would likely be non‑differential across groups.

In conclusion, the study demonstrates that (1) low lumbar BMD, particularly below ~84 mg/cc, is a significant risk factor for reoperation after non‑augmented lumbar fusion, and (2) PMMA‑augmented pedicle screw fixation provides comparable reoperation rates even in patients with severe osteoporosis. Opportunistic QCT derived from routine CT scans offers a practical, cost‑effective means of pre‑operative bone health assessment, potentially guiding surgical planning (e.g., decision to augment, construct length) and postoperative management (e.g., pharmacologic osteoporosis treatment). Prospective, larger‑scale studies with standardized reoperation criteria are warranted to confirm these findings and to refine BMD thresholds for clinical decision‑making.