Abstract: Objective  To investigate the relationship between serum low density lipoprotein (LDL) and the progression-free survival (PFS) of small cell lung cancer (SCLC) patients. Methods  A total of 271 SCLC patients admitted to Shandong Cancer Hospital from May 1, 2014 to October 31, 2018 were selected. These patients were divided into limited-stage group (n=126) and extensivestage group (n=145) according to Veteran′s Administration Lung Cancer Study Group (VALSG) evaluation standard. The correlation between the level of serum LDL before treatment and PFS was analyzed by Spearmen test in the two groups. After finding the cutoff value of LDL level by receiver operating characteristic curve (ROC) analysis, the relationship between LDL level before treatment and PFS was analyzed. According to the dynamic change of serum LDL during the treatment, extensive-stage patients were divided into four groups: normalized LDL group (n=25, patients whose LDL≤2.55 mmol/L and never increased until progression), increased LDL group (n=31, patients whose LDL≤2.55 mmol/L and increased at least once until progression), never-normalized LDL group (n=33, patients whose LDL>2.55 mmol/L and never normalized until progression), and decreased LDL group (n=56, patients whose LDL>2.55 mmol/L and decreased at least once until progression). Then the PFS among the four groups was compared. The survival curves were plotted by the Kaplan-Meier method and compared using the log-rank test. The significance of the independent variables for PFS in extensive-stage patients was analyzed using the Cox proportional hazards model. Results  The median PFS for the whole cohorts was 7.1 months (1.4-27.1 months). The median PFS for limited-stage patients and extensive-stage ones was 8.8 months and 6.1 months respectively, with a significant statistical difference (χ²=28.723, P＜0.001). LDL levels before treatment were negatively associated with PFS in all the patients (r=-0.234, P＜0.001) and extensive-stage group (r=-0.329, P＜0.001), but there was no statistical significance in limitedstage group (r=-0.119, P=0.183). The cutoff point of LDL was 2.55 mmol/L, with the highest value of sensitivity (55.56%) and specificity (81.69%) in the ROC analysis using PFS as an end point for extensive-stage patients. Patients in the low-LDL group （≤2.55 mmol/L, n=107） had relatively longer PFS compared to the ones in the high-LDL group （＞2.55 mmol/L, n=164） for whole cohorts (9.0 months vs. 6.2 months, χ²=16.064, P＜0.001). The serum LDL level before treatment showed a prognostic power mainly related to PFS within the extensive-stage cohort (χ²=21.419, P＜0.001), with no difference in the limited-stage cohort (χ²=2.718, P=0.099). In the extensive-stage cohort, the median PFS in the normalized LDL group, the increased LDL group, the nevernormalized LDL group and the decreased LDL group was 9.2, 6.5, 5.0 and 6.2 months respectively, and there was a significant difference in the PFS among the four groups (χ²=16.411, P＜0.001). Univariate analysis showed that the LDL>2.55 mmol/L before treatment (HR=0.436, 95%CI: 0.297-0.640, P＜0.001) and LDL>2.55 mmol/L and never normalized until progression (HR=2.215, 95%CI: 1.403-3.497, P＜0.001) were risk factors for PFS, LDL normal during treatment (HR=0.343, 95%CI: 0.190-0.618, P＜0.001) was the protective factor of PFS in extensive-stage patients. Multivariate Cox regression analysis showed that the LDL>2.55 mmol/L before treatment (HR=0.435, 95%CI: 0.300-0.632, P＜0.01) and LDL>2.55 mmol/L and never normalized until progression (HR=2.028, 95% CI: 1.386-2.966, P＜0.001) were independent risk factors for PFS, LDL normal during treatment (HR=0.318, 95%CI: 0.186-0.542, P＜0.001) was independent protective factors of PFS. Conclusion  The serum LDL level may be used as a potential predictive marker for PFS in extensive-stage SCLC patients subjected to the first-line chemotherapy. 

Key words: Lipoproteins, LDL, Small cell lung carcinoma, Treatment outcome, Forecasting, Progression-free survival