This retrospective cohort study analyzed medical record data from patients diagnosed with diffuse large B-cell lymphoma (DLBCL) at Dr. Cipto Mangunkusumo Hospital (RSCM) and the affiliated anatomical pathology laboratory between January 2014 and March 2021. The starting point (time zero) for survival analysis was defined as the date of initiation of first-cycle R-CHOP chemotherapy.

This retrospective study was approved by the Health Research Ethics Committee of the Faculty of Medicine, Universitas Indonesia and Cipto Mangunkusumo National Central General Hospital (reference number KET-133/UN2.F1/ETIK/PPM.00.02/2023). Given the retrospective nature of the study, written informed consent was obtained from surviving patients whenever feasible. For patients who had died prior to study initiation, the requirement for informed consent was waived by the ethics committee. Patient confidentiality was strictly maintained, and all data were anonymized prior to analysis.

Inclusion criteria were as follows: (a) age ≥18 years; (b) diagnosis of DLBCL according to the World Health Organization (WHO) classification; (c) receipt of an initial R-CHOP chemotherapy regimen consisting of 6–8 cycles, with or without radiotherapy; and (d) completion of chemotherapy with an average delay of no more than one week per cycle. Exclusion criteria included: (a) a history of indolent lymphoma or evidence of transformed DLBCL (e.g., transformation from follicular lymphoma or chronic lymphocytic leukemia), as these represent biologically distinct entities; (b) discordant or multiple histopathological diagnoses; (c) HIV infection; (d) primary central nervous system (CNS) DLBCL; (e) terminal comorbidities (e.g., end-stage renal failure, decompensated liver cirrhosis, or advanced heart failure); and (f) incomplete clinical or pathological data. Patient status was confirmed via telephone contact using contact information recorded in the medical charts. Baseline clinical characteristics collected included demographic data, number of extranodal sites, Ann Arbor stage, serum lactate dehydrogenase (LDH) levels, International Prognostic Index (IPI) score, and cell-of-origin classification based on the Hans algorithm. A flow diagram illustrating patient identification, exclusion criteria, and final cohort assembly is presented in Figure 1.

Absolute monocyte count (AMC) was obtained from peripheral blood samples analyzed using a Sysmex XN-1000 or XN-3000 hematology analyzer (Sysmex Corporation, Kobe, Japan).

Formalin-fixed, paraffin-embedded (FFPE) tissue blocks were sectioned at 3–4 μm and mounted on poly-L-lysine–coated slides. Sections were deparaffinized in xylene and rehydrated through a graded alcohol series. Antigen retrieval was performed using Tris–EDTA buffer (pH 9.0) in a decloaking chamber at 98°C for 10 minutes, followed by cooling and washing with phosphate-buffered saline (PBS). Endogenous peroxidase activity was blocked using SeyTek Super Block. Slides were incubated for 30 minutes with Bond™ ready-to-use primary antibodies against CD163 and CD8 (Biocare Medical, Netherlands). After washing, sections were treated with a Val Universal Linker secondary antibody and visualized using a high-contrast diaminobenzidine (DAB) substrate. Counterstaining was performed with Mayer’s hematoxylin, followed by bluing in lithium carbonate and dehydration through graded alcohol and xylene. Slides were coverslipped using entellan. Negative controls were included in each staining batch, while tonsil tissue served as a positive control. For quantitative analysis, five representative high-power fields (HPFs; 400× magnification, area 0.24 mm ²) were photographed using a Leica ICC50HD microscope with a camera attachment. Quantification of CD163-positive macrophages and CD8-positive lymphocytes was performed using ImageJ ^® software. CD163 expression was quantified as the number of positively stained macrophages per high-power field (cells/HPF). CD8 expression was quantified as the percentage of CD8-positive lymphocytes among total nucleated cells within each high-power field (%). Staining intensity was not incorporated into the analysis to reduce variability related to tissue fixation and processing. All evaluations were independently performed by two blinded investigators to ensure inter-observer reliability.

Event-free survival (EFS) was defined as the time from initiation of R-CHOP chemotherapy to the first occurrence of disease progression, relapse after initial response, initiation of second-line therapy, or death from any cause. Patients without an event were censored at the date of last follow-up.

Optimal cutoff values for AMC, CD163, and CD8 were determined using receiver operating characteristic (ROC) curve analysis based on two-year event-free survival. This approach was selected to provide clinically interpretable thresholds while minimizing overfitting in a retrospective cohort. Correlations between AMC, CD163, and CD8 were assessed using Spearman’s correlation test. The median follow-up duration, estimated using a reverse Kaplan–Meier approach, was 20 months. Event-free survival was analyzed using Kaplan–Meier curves and compared using the log-rank test. Univariate and multivariate Cox proportional hazards regression models were employed to estimate hazard ratios (HRs) and 95% confidence intervals (CIs). Variables with a p-value <0.05 in univariate analysis were included in the multivariate model. A two-sided p-value <0.05 was considered statistically significant. Although maximally selected rank statistics have been proposed for determining optimal cutoffs in time-to-event analyses, ROC-based thresholds were selected in this study to provide clinically interpretable values while minimizing model instability in a retrospective cohort with a limited sample size. Future prospective validation using alternative statistical approaches, including maximally selected rank statistics or time-dependent ROC methods, would be valuable. All statistical analyses were performed using SPSS version 20.0 (IBM Corp., Armonk, NY, USA). Missing data were handled using a complete-case analysis approach; variables with missing values were excluded from analyses requiring those parameters, and no data imputation was performed. All data were independently verified by two reviewers to ensure accuracy and completeness.

Between January 2014 and March 2021, a total of 108 patients diagnosed with diffuse large B-cell lymphoma (DLBCL) who met the inclusion and exclusion criteria were included in the study. Serum lactate dehydrogenase (LDH) data were missing in nine patients. Among all subjects, 56 (52%) were male, and 72 (66.7%) were younger than 60 years (median age: 53 years; range: 18–88 years). Early-stage disease (Ann Arbor stage I–II) was observed in 60 patients (55.6%), and 89 patients (82%) had 0–1 extranodal sites. Elevated LDH levels (>1× upper limit of normal) were present in 81 patients (81.8%), and 75 patients (75.8%) had a low-risk International Prognostic Index (IPI) score (0–2). Based on the Hans classification, the germinal center B-cell (GCB) subtype was identified in 33 patients (30.6%). During the two-year follow-up period, 58 patients (53.7%) experienced an event, while 46 patients (42.6%) remained event-free. Follow-up information was unavailable for four patients (3.7%). The median values of absolute monocyte count (AMC), CD163 expression, and CD8 expression were 619.2 cells/μL (interquartile range [IQR]: 455.0–759.45), 21.5 cells/HPF (IQR: 14.25–27.0 cells/HPF), and 23% (IQR: 14–39%), respectively. Detailed demographic and clinical characteristics are summarized in Table 1.

Receiver operating characteristic (ROC) curve analysis was performed to determine optimal cutoff values for AMC, CD163, and CD8 in predicting two-year event-free survival (EFS). The optimal cutoff for AMC was 631 cells/μL, yielding an area under the curve (AUC) of 0.923 (95% CI: 0.866–0.981), with a sensitivity of 82.8% and specificity of 90.0%. For CD163 expression, the optimal cutoff was 23 cells/HPF, with an AUC of 0.931 (95% CI: 0.880–0.983), sensitivity of 75.9%, and specificity of 96.0%. CD8 expression demonstrated an optimal cutoff value of 27.5%, with an AUC of 0.852 (95% CI: 0.773–0.930), sensitivity of 84.5%, and specificity of 84.0%.

ROC curves are presented in Figure 2. Patients were subsequently categorized into high- and low-expression groups according to these cutoff values ( Figure 3).

Spearman’s correlation analysis demonstrated a significant positive correlation between AMC and CD163 expression (r = 0.577, p < 0.001) and a significant negative correlation between AMC and CD8 expression (r = −0.599, p < 0.001). These findings suggest a potential association between peripheral monocyte levels and the immune composition of the tumor microenvironment. Correlation analyses are illustrated in Figure 4.

Patients with high AMC levels (≥631 cells/μL) had significantly poorer two-year EFS compared with those with lower AMC levels (HR = 9.82; 95% CI: 4.89–19.70; p < 0.001). Similarly, elevated CD163 expression (≥23 cells/HPF) was associated with inferior EFS outcomes (HR = 8.57; 95% CI: 4.58–16.05; p < 0.001). In contrast, patients with higher CD8 expression (≥27.5%) demonstrated significantly improved EFS (HR = 0.13; 95% CI: 0.06–0.26; p < 0.001). Subgroup analysis based on the Hans cell-of-origin classification showed no significant difference in EFS between the GCB and non-GCB subtypes (HR = 0.75; 95% CI: 0.42–1.35; p = 0.336). In contrast, a high IPI score was significantly associated with reduced EFS (HR = 3.06; 95% CI: 1.77–5.35; p < 0.001). Detailed survival analyses are presented in Table 2 and Figure 5. The median event-free survival time for the overall cohort was 13 months.

-

We hereby confirm that all the Figures and Tables in the manuscript are ours. Furthermore, any Figures and images, that are not ours, have been included with the necessary permission for re-publication, which is attached to the manuscript.