Objective To evaluate the value of the systemic immune-inflammatory index (SII), CEA, Cyfra21-1, and NSE in predicting and diagnosing bone metastasis of lung cancer.

Methods The clinical data of 618 patients with lung cancer were retrospectively analyzed. According to the bone metastasis at baseline, the data of the diagnosis group (patients with bone metastasis at baseline and patients without bone metastasis during follow-up) and the prediction group (patients with bone metastasis during follow-up and patients without bone metastasis during follow-up) were analyzed to determine the correlation between the above indicators and lung cancer bone metastasis.

Results Predictive group: SII≥850 and NSE≥58.64 ng/ml were independent risk factors and independent predictors for lung cancer bone metastasis. The AUC of the combined SII+NSE model was 0.662, with a sensitivity of 54.5% and a specificity of 74.5%; it was superior to the predictive value of single factor (95%CI: 0.596-0.728; P < 0.001). Diagnostic group: lung adenocarcinoma, SII≥951.6, CEA≥5.14 ng/ml, NSE≥20.15 ng/ml, and Cyfra21-1≥3.94 ng/ml were independent risk factors for bone metastasis in lung cancer patients (P < 0.05). The AUC of SII alone in the diagnosis of lung cancer bone metastasis was 0.754. The AUC of the SII+Cyfra21-1 combined model was 0.82 which was the largest, with a sensitivity of 74% and a specificity of 78.5%; it was superior to any univariate AUC (P < 0.05).

Conclusion The levels of SII, CEA, Cyfra21-1, and NSE in the bone metastasis group are significantly higher than those in the non-bone metastasis group. The predictive and diabnostic values would be improved further when SII combined with other single risk factors.