Serum low-density lipoprotein and low-density lipoprotein expression level at diagnosis are favorable prognostic factors in patients with small-cell lung cancer (SCLC)

This retrospective study involved data collection from SCLC patients between January 2004 to December 2011 at Sun Yat-Sen University Cancer Center (SYSUCC). All enrolled patients met the following criteria: (a) pathologically confirmed primary SCLC, (b) available clinical information, (c) normal liver function, and (d) detailed laboratory data, including cholesterol and LDL at diagnosis. In both groups, patients were recruited with lipid metabolism-related diseases, or currently treated for concomitant diseases that would influence serum lipids (i.e., diabetes, hyperlipidemia, or metabolic syndrome), patients with liver disease, or other types of cancer. A total 601 eligible patients were enrolled into the study. Among them, 198 cases have sufficient tumor specimens for immunohistochemistry (IHC). All patients were staged according to the Veterans Administration Lung Study Group (VALSG) staging system. Complete clinical information of all patients (i.e., demographics, performance status, treatments and laboratory tests) was recorded. Smokers were defined as patients who had more than 100 cigarettes. The study was approved by the Institutional Review Board of SYSUCC and written informed consent was obtained for each patient prior to sample collection.

Most patients received four cycles of platinum plus etoposide as chemotherapy, and some patients also were subsequently treated with prophylactic cranial irradiation (PCI). Several patients underwent thorax radiotherapy (TRT) in accordance with chemotherapy.

We performed IHC staining to evaluate the expression of LDLR in SCLC patients. Sections (thickness, 3–4 μm) were deparaffinized and rehydrated. For antigen retrieval, the slides were soaked in ethylene diamine tetraacetic acid (EDTA) and Aantigen Rretrieval Ssolution (3000 ml, pH 8.0), followed by heating in a pressure cooker for 12 min. Treated sections were then cooled to room temperature prior to immersing in distilled water for 2 min. To block the endogenous peroxidase activity and reduce non-specific assimilation, sections were treated with 3% H2O2 for 8 min, and further incubated in 5% bovine serum albumin for 30 min. Anti-LDLR (mouse LDLR antibody; R&D Systems; American) (1:400 dilution) was then added and incubated at 4 °C for 24 h. After washing with phosphate-buffered saline (PBS) for three cycles of 2 min, slides were incubated with secondary antibody (PV-9003 goat kit; ZSGB-Bio, Beijing, China) at 37 °C for 30 min. Afterwards, slides were washed with PBS thrice again. 3, 3′- diamino benzidine was applied for dyeing and hematoxylin was used to counterstain the sections. All sections were independently reviewed by two pathologists. Semi-quantitative scoring was used to evaluate immune reaction [25]. An IHC score, called HSCORE, was then applied to each sample based on the intensity of staining and the percentage of positive tumor cells. The HSCORE was calculated as following: HSCORE = Ʃ (I × PC). “I “means the intensity of staining and “PC” represents the percentage of positive tumor cells.

All the patients were carefully followed. Patients were evaluated every 2 months after completion of anti-tumor therapy. Routine follow-up examination was performed by computed tomography (CT) scan or/and Magnetic Resonance Imaging (MRI), including chest radiograph, abdominal ultrasonography and brain when clinically indicated. Anti-tumor response was assessed by radiologists according to the Response Evaluation Criteria in Solid Tumors (RECIST, version 1.1). Overall survival (OS) was defined as the months from the diagnosis to the death for any cause or last follow-up. Progression-free survival (PFS) was defined as the months from the diagnosis to the earliest occurrence of disease progression or death for any reasons. Patients who were alive at the time of last follow-up or lost to follow-up were censored. The last follow-up date was determine at May 31, 2015.

The patient baseline characteristics are presented in Table 1. A total of 601 patients with SCLC were enrolled in the study, with a median age of 60 years (range, 19–82 years). The majority of the patients were males (n = 529, 88%) and smokers (n = 505, 84.0%), and had a PS of 0–1 (n = 550, 91.7%). Among them, 254 (42.3%) patients had distant metastasis at the time of diagnosis and 347 (57.7%) patients were at limited stage. Most of the patients had prior treatment of two to four cycles of etoposide-based chemotherapy, while 22.1% (n = 133) patients received PCI and 37.8% (n = 227) had TRT. At last follow-up date, 433 (72%) patients had died. The median follow-up time was 31.75 months (range, 3.32 months to 117.41 months).

Using X-tile [26], we determined that the optimal cutoff for serum LDL in assessing OS is 2.14 and 3.36. Patients were divided into three groups based on the cutoff value of LDL: (1) low-LDL group (LDL level ≤ 2.14 mmol/L, n = 66, 11.0%) (2) intermediate-LDL group (3.36 mmol/L < LDL level ≥ 2.14 mmol/L, n = 282, 46.9%) and (3) High-LDL group (LDL level > 3.36 mmol/L, n = 253, 42.1%) (Additional file 1: Figure S1).

The clinicopathological characteristics of SCLC patients based on LDL levels are presented in Table 2. More patients had low levels of serum LDL in the etoposide-based chemotherapy group compared with those in the non-etoposide-based chemotherapy group (P = 0.011, Table 2). However, the level of LDL was not significant associated with age (P = 0.648), gender (P = 0.918), PS (P = 0.119), smoking status (P = 0.411), and disease stage (P = 0.189) (Table 2).

Based on the results by X-tile, the 198 patients who had sufficient tumor samples for IHC were grouped as followed: (1) low-LDLR group (HSCORE ≤ 60, n = 145, 73.2%, Fig. 1a), and (2) high-LDLR group (HSCORE >60, n = 53, 26.8%, Fig. 1b) (Additional file 1: Figure S1).

Among the 198 patients, 175 ones were males, ones were smokers, and ones had a PS of 0–1. The relationship between LDLR and clinical features have been analyzed. There were no obvious correlation of LDLR to gender (P = 0.565), PS (P = 0.118), and smoking status (P = 0.069).

The median OS for the 601 eligible patients was 15.43 months (range, 0.03–123.43 months). The median PFS for the entire cohort was 5.32 months (range, 0.03–71.79 months). A total of 66, 282, and 253 patients were categorized as low-LDL, intermediate-LDL, and high-LDL groups. Compared with the low-LDL group, patients with intermediate-LDL or high-LDL had lower survival outcome (low-LDL vs. intermediate-LDL vs. high-LDL, 29.27 vs. 16.70 vs. 17.23 months, respectively; P = 0.003) (Fig. 2a). When stratified by cancer stage, we found that LDL also showed a prognostic power in limited stage (P = 0.01, Fig. 2b). Moreover, baseline serum LDL value also had distinct significance in predicting PFS (P = 0.037, Fig. 2c).

To provide a significant control and a point of reference for LDLR expression, we also study the immune-staining of healthy lung tissue. The image shows that the expression of LDLR in healthy lung tissue is very low (Additional file 2: Figure S2). Moreover, patients with a lower expression of LDLR demonstrated significantly better OS (19.94 vs. 11.27 months, respectively; P = 0.003, Fig. 2d).

Other than LDL and LDLR, PS score (P < 0.001), smoking status (P < 0.001), and disease stage (P < 0.001) were also significantly associated with OS. Patients who received TRT (P < 0.001) or PCT (P = 0.001) also were associated with better OS (Fig. 3). However, there were no distinct associations between OS and gender (P = 0.438) and age (P = 0.424) (Table 1). In addition, patients with a lower PS score and in limited stage demonstrated significantly better PFS (P = 0.002 and P < 0.001, respectively).

Multivariate analyses, was performed to test for correlation among the different prognostic variables. We observed that higher LDL level was a significantly independent prognostic factor for poorer OS (P = 0.019, Table 3). Patients in intermediate LDL group were estimated to have 1.42-fold higher risk of death compared with those in the low LDL group (HR, 1.42; 95% CI: 1.08–2.03; P = 0.015). Patients with a LDL level of >3.36 had 1.64-fold higher risk of death than those in the low LDL group. Similarly, the multivariate analyses demonstrated that LDLR expression independently predicted OS in patients with SCLC (P = 0.027, Table 3). Compared with patients with a HSCORE ≤ 60, those with a HSCORE >60 had a 1.538 higher risk of death. Moreover, PS score (P < 0.001), cancer stage (P < 0.001), PCI (P = 0.011), and TRT (P = 0.007) were also independent predictors of survival outcome in patients with SCLC (Table 3).