The application of lung immune prognostic index in predicting the prognosis of 302 STS patients

Abstract

Background: Soft tissue sarcoma (STS) are heterogeneous and rare tumors, and few studies have explored predicting the prognosis of patients with STS. The Lung Immune Prognostic Index (LIPI), calculated based on baseline serum lactate dehydrogenase (LDH) and the derived neutrophils/(leukocytes minus neutrophils) ratio (dNLR), was considered effective in predicting the prognosis of patients with pulmonary cancer and other malignancies. However, the efficacy of the LIPI in predicting the prognosis of patients with STS remains unclear.

Methods: This study retrospectively reviewed patients with STS admitted to our center from January 2016 to January 2021. Their hematological and clinical characteristics were collected and analyzed to construct the LIPI specific to STS. The correlations between various predictive factors and overall survival (OS) were examined using Kaplan-Meier and Cox regression analyses. Independent risk factors for OS were identified using univariate and multivariate analyses. Finally, a LIPI nomogram model for STS was established.

Results: This study enrolled 302 patients with STS, of which 87 (28.9%), 162 (53.6%), and 53 (17.5%) were classified into three LIPI-based categories: good, moderate, and poor, respectively (P < 0.0001). The time-dependent operator curve showed that the LIPI had better prognostic predictive ability than other hematological and clinical characteristics. Univariate and multivariate analyses identified the Fédération Nationale des Centres de Lutte Contre le Cancer grade (FNCLCC/G), tumor size, and LIPI as independent risk factors. Finally, a nomogram was constructed by integrating the significant prognostic factors. Its C-index was 0.72, and the calibration curve indicated that it could accurately predict the three- and five-year OS of patients with STS. The decision and clinical impact curves also indicated that implementing this LIPI-nomogram could significantly benefit patients with STS.

Conclusion: This study explored the efficacy of the LIPI in predicting the prognosis of 302 patients with STS, classifying them into three categories to evaluate the prognosis. It also reconstructed a LIPI-based nomogram to assist clinicians in predicting the three- and five-year OS of patients with STS, potentially enabling timely intervention and customized management.

Keywords: derived neutrophil to lymphocyte ratio (dNLR); lactate dehydrogenase (LDH); lung immune prognostic index (LIPI); prognosis; soft tissue sarcoma.

Figure 1

The flowchart for patient selection and study design.

Figure 2

Conducting ROC analysis for various hematologic biomarkers. (A–D) The AUC and optimal cutoff values of NLR, PLR, dNLR, and LDH are as follows. Sensitivity is represented on the vertical axis, while 1-specificity is depicted on the horizontal axis.

Figure 3

(A–E) Various hematological indexes of 302 patients with STS are reflected by KM survival curves.

Figure 4

(A) Time-dependent ROC curves illustrate the variances in predictive capabilities of different hematologic markers. (B) The predictive abilities of STS independent prognostic factors are depicted by time-dependent ROC curves. Changes in predictive capabilities are reflected by time-dependent ROC curves, where a higher AUC value indicates superior predictive performance of STS.

Figure 5

(A) Conducting univariate analysis for clinical characteristics and hematological biomarkers. (B) Conducting multivariate analysis for significant clinical characters and hematological biomarkers.

Figure 6

The STS overall survival nomogram of STS was constructed and validated. (A) LIPI, FNCLCC/G, and size are combined to construct the nomogram, and the total score of the nomogram was the sum of the scores of each covariate. (B–D) The calibration curve, decision curve analysis, and clinical impact curve verified the nomogram.