Background
Diffuse large B-cell Lymphoma (DLBCL) is the most common type of aggressive B-cell non-Hodgkin’s lymphoma worldwide. Regarding the cell of origin, DLBCL can be classified as germinal center B-cell-like (GCB) and non-GCB cells by immunophenotypic detection. It can also be divided into activated B-cell-like, GCB, unclassified, or type III based on gene expression profiling. Regardless of cell origin, the standard first-line treatment for DLBCL is R-CHOP, which comprises the anti-CD20 monoclonal antibody rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone. However, 40–50% of patients with DLBCL suffer from refractory disease or relapse throughout the administration of standard first-line treatment or follow-up [
1‐
4]. Therefore, exploring the factors related to DLBCL pathogenesis may be helpful in the treatment and prognosis of patients.
Metabolic reprogramming is an important hallmark of tumor cells. Abnormal lipid metabolism has been reported in several types of tumors, which affects development and progression by regulating various oncogenic signal pathways [
5‐
7]. Few studies have demonstrated that abnormal lipid characterization is associated with R-CHOP resistance in a mouse model of DLBCL and that serum free fatty acids, high-density lipoproteins (HDL), low-density lipoproteins, and triglycerides are associated with the prognosis of patients with DLBCL [
8‐
12].
Apolipoprotein A1 (ApoA1), the main component of HDL, has been extensively studied in the context of atherosclerosis, thrombotic diseases, diabetes, and nervous system diseases. The findings of these studies showed that ApoA1 could bind to ATP-binding cassette transporter (ABCA1) of macrophages and endothelial cells, trigger cholesterol efflux from these cells, then reduce the production of inflammatory factors in these cells and the degeneration of macrophages, leading to protection of arterial wall. It could improve the integrity of endothelial barrier, inhibit the activity of von Willebrand factor (vWF) activity and increase the degradation of procoagulant factors, then reduce thrombosis. It could also bind to ABCA1 of β-cell, inhibit its’ apoptosis and increase insulin secretion, then improve glycaemic control. Furthermore, it could cross the blood-brain barrier, reduce amyloid beta (Aβ) deposition and levels, as well as neuroinflammation, exhibiting neuroprotective functions [
13‐
15]. In recent years, an increasing number of studies on ApoA1 in cancers have been conducted [
16‐
26]. However, the possible role of APOA1 in the pathogenesis and prognosis of DLBCL has been poorly studied [
11,
12]. Therefore, we retrospectively analyzed the data of newly diagnosed patients with DLBCL at our center to explore the prognostic value of pretreatment serum ApoA1 levels.
Methods
Study population
We studied all consecutive patients with newly diagnosed DLBCL who were admitted to the Fujian Medical University Union Hospital between January 1, 2011, and December 31, 2021. All patients were diagnosed according to the World Health Organization (WHO) 4th revision classification of tumours of haematopoietic and lymphoid tissues. Newly diagnosed patients with DLBCL who were ≥ 14 years and received ≥ 4 cycles of chemotherapy or immunochemotherapy were included in this study. Patients diagnosed with primary mediastinal lymphoma, primary central nervous system lymphoma, human immunodeficiency virus infection, diabetes mellitus, or other malignancies were excluded from the study. Demographic information, clinical characteristics, and laboratory parameters, including age, sex, B symptoms, serum lactate dehydrogenase (LDH) levels, Eastern Cooperative Oncology Group (ECOG) score, Ann Arbor stage, extranodal disease sites, International Prognostic Index (IPI) score, Hans classification, serum ApoA1 levels, and therapeutic regimens, were obtained from medical records.
The same number of healthy individuals without metabolic and neoplastic diseases who presented in the aforementioned period were sex- and age-matched to the DLBCL group and assigned to the control group.
Approval for this study was obtained from the Ethics Committee of Fujian Medical University Union Hospital. As this study was a retrospective data analysis and did not affect patients’ treatments, written informed consent was not sought. This study was conducted in accordance with the Declaration of Helsinki.
Sample collection and detection
Peripheral blood samples were collected from all patients within a week of primary therapy. Peripheral blood samples were collected from 949 healthy individuals who were included in the control group. All samples were collected from the patients and healthy individuals after overnight fasting.
Serum ApoA1 levels were measured by an immunoturbidimetry assay using a Roche cobas8000 automatic biochemical analyzer. Serum ApoA1 detection kit was purchased from Roche (Basel, Switzerland).
Treatment and follow-up
All newly diagnosed patients with DLBCL included in this study had received ≥ 4 cycles of chemotherapy with or without a history of tissue biopsy or surgical excision. The therapeutic regimen comprised cyclophosphamide, doxorubicin, epirubicin, vincristine, and prednisone with or without rituximab. Patients who received fewer than three cycles of rituximab were assigned to the CHOP-like group; otherwise, patients were assigned to the R-CHOP-like group. The treatment response was evaluated according to the Revised Response Criteria for Malignant Lymphoma [
27]. For disease progression or relapse, patients were treated with later-line regimens, as recommended by the NCCN guidelines [
28‐
31].
Overall survival (OS) was calculated from the date on which the patient started treatment to the date of death from any cause or last follow-up. Progression-free survival (PFS) was calculated from the date on which the patient started treatment to the date of disease progression after first-line chemotherapy or immunochemotherapy, relapse, death, or last follow-up. The last follow-up date was February 10, 2023. Follow-up data were obtained from the clinical records or via telephone interviews.
Statistical analyses
All statistical analyses were performed using SPSS software version 19.0 (IBM Corp., Armonk, NY, USA) and GraphPad Prism version 8.0 (GraphPad Corp., Boston, USA). The receiver operating characteristic (ROC) curve was used to determine the optimal cutoff value for serum ApoA1 levels. Categorical and continuous variables were compared using the χ2-test and t-test, respectively. Time-to-event data were analyzed using the Kaplan-Meier method. The log-rank test was used to compare the survival times of different groups. The Cox proportional hazards model was used for univariate analysis of the potential predictors of survival. The Cox regression model was used for the multivariate analysis of variables identified as significant prognostic factors in the univariate analysis. Statistical significance was defined as a two-sided P-value < 0.05.
Discussion
Conventionally, the heterogeneity of immunology, cytogenetics, and molecular biology exists in DLBCL tissue, leading to heterogeneity in survival among patients with DLBCL. Currently, clinical studies have demonstrated that 40–50% of patients suffer from refractory or relapsed DLBCL after R-CHOP treatment [
1‐
4]. Some studies have shown that intensified immunochemotherapy or addition of novel drugs to the R-CHOP regimen can improve the treatment response and prognosis [
32‐
35]. Therefore, risk stratification of newly diagnosed patients with DLBCL is critical for the choice of treatment and, thus, affects outcomes. Thus far, indices for risk stratification of DLBCL include the IPI score and its derivatives, revised IPI and National Comprehensive Cancer Network IPI (R-IPI and NCCN-IPI). These indices comprise clinical factors including age, Ann Arbor stage, extranodal disease, ECOG score, and serum LDH level. However, none of these indices can efficiently distinguish high-risk patients from others [
36]. Hence, it is necessary to identify other factors related to the development and progression of DLBCL, determine their role in the prognosis of patients with this disease, and establish useful prognostic indices.
Lipid metabolism-related genes and products have been reported to be involved in tumorigenesis and serve as biomarkers for tumor diagnosis and prognosis [
5,
14,
37‐
39]. ApoA1 is a member of the apolipoprotein family that participates in reverse cholesterol transport and lipid metabolism. Basic studies have revealed that APOA1 plays a key role in the development and progression of solid tumors. In colorectal cancer, the expression of ApoA1 in tumor cells is associated with different metastatic tropisms [
26]. In patients with squamous cervical cancer, APOA1 overexpression can reduce the sensitivity of cervical squamous cells to carboplatin by regulating the expression of STAT1, CD81, C3, and TOP2A [
19]. In osteosarcoma, circ_0088212 exhibits a cancer-promoting role by regulating the miR-520 h/APOA1 axis [
40]. In pancreatic ductal adenocarcinoma, ubiquitination and degradation of ApoA1 mediated by TRIM15 can promote the invasion and metastasis of pancreatic cancer cells. The APOA1-LDLR axis also participates in the metastasis of pancreatic cancer cells by regulating lipid metabolism in tumor cells. Moreover, APOA1 overexpression can reduce tumor growth in vivo [
23,
41]. Clinical studies have shown that abnormal ApoA1 levels in the serum or tissues of patients are associated with the prognosis of many types of tumors. In patients with breast cancer, the frequency of APOA1 copy number loss in tumor cells is negatively associated with serum ApoA1 levels. Lower ApoA1 levels were associated with a higher incidence of intraocular metastasis [
20]. In patients with advanced non-small cell lung cancer, serum ApoA1 levels are associated with the frequency of EGFR T790M mutations and treatment response to EGFR tyrosine kinase inhibitors [
21]. In patients with kidney clear cell carcinoma (KIRC), the expression levels of APOA1 mRNA in KIRC tissues were lower than those in para-cancerous and normal kidney tissues, and high expression of APOA1 mRNA in tissues was correlated with worse OS and disease-free survival. Moreover, the preoperative serum ApoA1 levels in patients with KIRC were significantly lower than those in healthy controls, and patients with KIRC with low serum ApoA1 levels had worse OS [
24]. Serum ApoA1 level is an independent prognostic factor for OS in patients with CRC and advanced intrahepatic cholangiocarcinoma treated with PD-1 [
22,
25]. Recently, Wang et al. found that low serum ApoA1 levels were associated with the PFS and OS of patients with DLBCL but were not an independent prognostic factor for survival in a cohort of 307 newly diagnosed patients with DLBCL [
11]. While Yu et al. found that low serum ApoA1 levels were independent poor prognostic factor for OS and PFS in a cohort of 105 patients with DLBCL [
12]. Since research on serum ApoA1 levels in newly diagnosed patients with DLBCL and its possible prognostic significance is scarce, we sought to investigate these matters in a larger DLBCL population.
In this study, we found that the serum ApoA1 levels were lower in patients with DLBCL than in healthy individuals. Similar result was also found in the study by Yu et al. [
12]. Thereafter, we identified the optimal cutoff value for serum ApoA1 level, accordingly stratified patients into the low ApoA1 and high ApoA1 groups, and analyzed the relationships between serum ApoA1 levels and clinical and laboratory parameters in these two groups. We found that patients with DLBCL with low serum ApoA1 levels had a higher percentage of male, more B symptoms, higher levels of LDH, poorer performance status, higher percentage of advanced stage, more cases in which there was more than one extranodal site, higher IPI, and a higher percentage of non-GCB subtypes. Interestingly, we also found that the incidence of relapse or refractory disease in patients with DLBCL with low serum ApoA1 levels was significantly higher than that in patients with DLBCL with high serum ApoA1 levels. Moreover, Kaplan-Meier analysis demonstrated that low ApoA1 levels were associated with shorter PFS in patients with DLBCL, which was similar to the results of two previous study [
11,
12]. Since the occurrence of relapse or refractory disease was associated with PFS, these results suggest that serum ApoA1 level may be a possible biomarker of treatment response and helpful for the choice of therapeutic regimens. Furthermore, multivariate analysis demonstrated that an ECOG score ≥ 2, LDH > ULN, non-GCB subtype, and absence of rituximab were independent poor prognostic factors for OS and PFS in patients with newly diagnosed DLBCL. Ann Arbor Stage III–IV and more than one extranodal site were independent poor prognostic factors for PFS, while low serum ApoA1 levels were independent poor prognostic factors for OS. This result is inconsistent with that found by Yu et al. [
12]. Then we compared the clinical characteristics of patients with DLBCL between our cohort and Yu et al. cohort. We found that the sample size of Yu et al. cohort was so small. Furthermore, the percentage of patients with IPI > 2, or Ann Arbor Stage III–IV, or extranodal disease > 1, or ECOG ≥ 2, or LDH > ULN in their cohort was lower than us (26.7% vs. 40.0%, 58.1% vs. 66.0%, 14.3% vs. 34.0%, 6.7% vs. 21.1% and 40% vs. 49.7%), respectively. We think that the differences in clinical characteristics between these two cohort account for these inconsistent results, too.
Nevertheless, this study had some limitations. The selection bias from the retrospective analysis limited the efficacy of our results. Because the detection methods of ApoA1 in multiple centers were different, the optimal cut-off value of serum ApoA1 from our study may not be applicable for prognostic analysis in other centers [
14,
20]. Since there were too many missing data on cytogenetic or molecular biological information in our cohort, we were unable to reveal the correlation between serum ApoA1 levels and the biological tumor characteristics of patients with DLBCL. Although serum ApoA1 levels are associated with treatment response in patients with DLBCL, the role of the APOA1 gene in the development and progression of DLBCL and its underlying mechanisms have not been confirmed. Previous studies demonstrated that ApoA1 has been shown to be involved in chemoresistance and metastasis in some types of solid tumors, and ApoA1 mimetic peptides had been confirmed to have anti-tumor effects [
15]. It suggests that ApoA1 may be a potential target for DLBCL treatment. Therefore, prospective multicenter studies with comprehensive cytogenetic or molecular biological information on patients and related basic research are needed to elucidate the role of serum ApoA1 in the development and progression of DLBCL and its prognostic predictive efficacy.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.