Background
Diffuse large B-cell lymphoma (DLBCL) is the most common adult non-Hodgkin lymphoma (NHL), accounting for 30–40%. Although aggressive, it can be cured in 60–70% of cases after first-line immunochemotherapy. Nevertheless, 30–40% of cases will experience recurrence or refractory disease after initial response, which will dramatically reduce their survival time. These patients remain a challenging therapeutic problem, and more targeted and personalized approach is an important goal for them, which is still being explored [
1].
In the past decades, the view that tumors mainly consist of tumor cells has changed, and tumors are now thought to have organoid structures. In addition to tumor cells, there are many other tumor-infiltrating stromal cells, such as fibroblasts, immune cells, vascular endothelial cells and mesenchymal stem cells, which interact closely with tumor cells, constituting the tumor microenvironment (TME) [
2]. TME influences initiation, growth and metastasis of tumors, including DLBCL.
Tumor-associated macrophages (TAMs), as the intrinsic cellular components of TME, involve in tumor proliferation, invasion, angiogenesis, metastasis and suppression of anti-tumor immunity in many tumors [
3]. In general, macrophages (MPs) often have two kinds of phenotypes because of different activation states. One is “classically” activated M1 phenotype with antitumor activity, the other is “alternatively” activated M2 phenotype with tumor-promoting activity [
4]. So far, the clinical prognostic significance of TAMs in DLBCL remains controversial mainly due to difference in the method used to evaluate MPs as well as types of treatment given [
5‐
13].
Another known prognostic marker in DLBCL is the peripheral absolute monocyte count (AMC), and elevated AMC levels have been reported to correlate with poor prognosis in DLBCL patients [
14‐
20]. It is widely known that the majority of TAMs are derived from peripheral monocytes, however, the correlation between TAMs and AMC has not yet been elucidated in DLBCL. Therefore, in this study, we evaluated the predictive value of TAMs and clarified the correlation between TAMs and AMC in DLBCL patients.
Discussion
The present study evaluated the clinical prognostic implications of TAMs in DLBCL, as well as the association with AMC. We used 2 markers to identify TAMs, CD68 and CD163, and found that high CD68 or CD163 expression was correlated with clinico- pathological characteristics, high CD163 expression was an adverse predictor for both OS and PFS. Patients with high CD68 or high CD163 expression had significantly poorer OS and PFS than those with low CD68 or low CD163 expression, respectively, even in the rituximab era. Moreover, high-risk patients could be further identified by the expression of CD68 or CD163, especially in those classified as low/intermediate risk by IPI. Furthermore, the significant positive correlation was also detected between TAMs and AMC.
A recognized hallmark of cancer is the non-resolving inflammation in TME. In addition to tumor cells, there are different stromal cells influencing the tumorigenesis, growth, metastasis and angiogenesis of tumors in TME, such as fibroblasts, leukocytes (including myeloid lineage cells and lymphoid lineage cells), vascular endothelial cells and mesenchymal stem cells. TAMs, generally thought to be more similar to M2- polarized MPs, are often prominent stromal cells that coordinate various factors in TME, and are also known as alternatively activated MPs, which are activated by helper T cell 2 (Th2) cytokines, such as interleukin (IL)-4, IL-10 and IL-13 [
23]. In recent studies, TAMs in TME were found to correlate with clinical prognosis in hematologic neoplasms, particularly in Hodgkin’s lymphoma (HL) [
24]. Nevertheless, its significance in DLBCL has thus far been controversial. The number of CD68 + cells did not exhibit significant prognostic value [
13], and was found to have no significant relation with OS or PFS [
5], whereas other researches showed that higher CD68 + cells tended to predict poorer prognosis [
6,
8,
9] and worse treatment response [
6,
11]. But recently, Nam et al. [
8] and Riihijärvi S et al. [
9] suggested that higher CD68 expression was correlated with better prognosis in DLBCL received rituximab in addition to other chemotherapy such as CHOP, and with poorer outcome when rituximab was not given. Our study results are in line with these findings [
7,
8], suggesting that high CD68 expression was correlated with clinicopathological characteristics and had significantly poor OS and PFS, even in the rituximab era. In addition, it is generally believed that high CD163 expression predicts poor prognosis in DLBCL patients [
7,
8,
10,
12,
13], and our results are consistent with those previous studies. The clinical prognostic significance of TAMs in DLBCL remains controversial possibly due to difference in clinical case-study method, patient populations, instrument (such as field of view) and method used to evaluate MPs as well as types of treatment given. In summary, TAMs may play an important role in promoting tumor growth, invasion and metastasis in TME of DLBCL, and CD163, as considered to be a marker relatively specific to M2 TAMs may serve as a better predictor than the pan-macrophage marker, CD68. There have been several studies on the role of TAMs in DLBCL. TAMs remodel extracellular matrix (ECM) remodeling and promote angiogenesis through producing matrix metalloprotein-9 (MMP-9), legumain, and vascular endothelial growth factor (VEGF) [
25,
26]. Reactive oxygen species (ROS) from TAMs contributes to DLBCL progression and drug resistance by increasing the expression of CD44 [
27]. TAMs express programmed death-1 ligand (PD-L1), which induces the apoptosis of T cells, involving immunosuppression and immune evasion [
28,
29]. DLBCL cells can avoid phagocytosis through CD47/signal regulatory protein α (SIRPα) on TAMs pathway [
30].
Monocytes and their progeny, promote tumor growth and angiogenesis, and involve in the suppression of host anti-tumor immunity. So more and more studies have found that elevation of peripheral monocytes is an adverse prognostic factor in many solid tumors [
31‐
34], including DLBCL [
14‐
20]. We have previously reported that peripheral AMC at diagnosis predicted outcome for DLBCL patients who received standard first-line regimens [
14], elevation of AMC at the first relapse was an adverse prognostic factor for survive in relapsed/refractory DLBCL [
15], and increased AMC tested during follow-up after standard first-line regimens was a risk factor for predicting recurrence of DLBCL [
16]. Until now, few studies [
34‐
37] have demonstrated that peripheral AMC significantly correlated with TAMs. However, to our knowledge, in DLBCL the correlation between AMC and TAMs has not been clarified. In our study, AMC was significantly positively correlated with TAMs in DLBCL patients. So, why peripheral blood monocytosis indicates poor prognosis in DLBCL, we deduce that monocytes differentiated into MPs after recruitment from peripheral blood to TME, then polarized into M2 phenotype thereby causing the progression of disease.
There are some limitations in the study. Firstly, as a retrospective study, patient selection might have been biased, and other unrecognized bias might have influenced the results. Secondly, the amount of cases is relatively small to prove conclusions. Thirdly, we used ROC curves and AUC to determine the best TAMs (CD68 + cells and CD163 + cells) cutoff value in our study, however, the cut-point for TAMs need investigation in other independent cohort.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.