Dynamic inflammatory biomarkers as prognostic indicators in advanced breast cancer patients receiving PD-1 inhibitors

Abstract

Background:

While programmed cell death protein 1 (PD-1) inhibitors benefit breast cancer patients, simple predictors of their efficacy are lacking.

Objectives:

This study aimed to evaluate inflammatory markers as prognostic indicators for advanced breast cancer (ABC) patients receiving immunotherapy.

Design:

This is a single-center retrospective study of ABC patients treated with PD-1 inhibitors between January 2016 and June 2022.

Methods:

Clinicopathological parameters, tumor-infiltrating lymphocytes (TILs) levels, and inflammatory markers—C-reactive protein (CRP), neutrophil-to-lymphocyte ratio (NLR), and lymphocyte-to-monocyte ratio (LMR)—were collected from 116 ABC patients at baseline and after three cycles of immunotherapy (CRP3, NLR3, LMR3). Optimal cut-offs were defined using R software and associations with progression-free survival (PFS), overall survival (OS), and objective response rate were assessed.

Results:

Low baseline CRP, low NLR, and high LMR were significantly associated with improved PFS and OS (all p < 0.05) by Kaplan–Meier analysis. On multivariate analysis, low NLR3 and a reduced CRP3/CRP ratio independently predicted prolonged PFS (p < 0.05). Patients with low CRP, reduced CRP3/CRP ratio, low NLR, and low NLR3 achieved better OS (p < 0.05). Combining inflammatory markers with TIL status improved prognostic classification: TIL-deficient patients with high NLR3 (median PFS (mPFS): 2.89 months) or high CRP3/CRP ratio (mPFS: 1.90 months) had the poorest survival (p < 0.05).

Conclusion:

CRP and NLR are promising biomarkers for predicting outcomes of PD-1 inhibitor therapy in ABC, with potential utility in individualizing immunotherapy strategies.

Keywords

advanced breast cancer C-reactive protein immunotherapy neutrophil-to-lymphocyte ratio prognostic biomarker tumor-infiltrating lymphocytes

Background

Programmed cell death protein 1 (PD-1) inhibitors, as representative immune checkpoint inhibitors (ICIs), have shown potential for the treatment of breast cancer.^1–5 However, breast cancer’s heterogeneity and “cold tumor” characteristics lead to variable immunotherapy responses, highlighting the need for predictive biomarkers. PD-L1 expression, tumor-infiltrating lymphocytes (TILs), and tumor mutation burden are potential biomarkers,^6–8 their clinical application is limited by challenges such as PD-L1 heterogeneity, technical complexity, and high cost. Therefore, accurate, simple, and accessible biomarkers are crucial for identifying patients most likely to benefit from immunotherapy.

Inflammation plays a pivotal role in tumor development, progression, and the modulation of the tumor immune microenvironment, thereby influencing responses to PD-1/PD-L1 blockade.^9–12 C-reactive protein (CRP) level, the neutrophil-to-lymphocyte ratio (NLR), and the lymphocyte-to-monocyte ratio (LMR) are commonly used inflammatory markers. CRP is an acute-phase protein produced by hepatocytes in response to circulating interleukin-6 (IL-6), a central mediator of the IL-6–JAK–STAT3 inflammatory pathway associated with tumor-related immune suppression.^13–15 Because CRP production is largely IL-6-driven, circulating CRP levels may serve as a readily measurable surrogate of this inflammatory signaling axis. Consequently, changes in CRP during PD-1 inhibitor therapy may reflect systemic immune modulation. While some studies link low baseline CRP to better immunotherapy response¹⁶ and high NLR to poorer prognosis in various cancers,^17–22 others report conflicting results regarding NLR and progression-free survival (PFS).^23,24 Additionally, an increased LMR has been associated with improved efficacy of nivolumab in advanced lung cancer.²⁵ However, most studies evaluating inflammatory markers in breast cancer have several limitations, including small sample sizes, a focus on a single subtype of breast cancer and/or a single inflammatory marker, or a lack of analyses of overall survival (OS). The specific roles of various inflammatory markers in immunotherapy for advanced breast cancer (ABC) require further in-depth investigation.

TILs have a well-established association with improved treatment response and prolonged survival after ICI therapy in triple-negative breast cancer (TNBC).^26,27 Nevertheless, as TILs evaluation is restricted to the local tumor microenvironment, it may not comprehensively reflect a patient’s overall systemic immune status. Consequently, the potential for systemic inflammatory biomarkers—such as the NLR to synergize with TILs in augmenting predictive accuracy within breast cancer immunotherapy warrants further investigation. To address these limitations, we retrospectively evaluated CRP, NLR, LMR, and platelet-to-lymphocyte ratio (PLR) in PD-1 treated ABC patients and integrated their dynamic changes with TILs to refine prognostic stratification.

Methods

Patients

We retrospectively reviewed data from adult patients (⩾18 years) with ABC treated with PD‑1 inhibitors between January 2016 and June 2022 at Sun Yat‑sen University Cancer Center. Patients with missing efficacy evaluations were excluded.

Data collection

Clinical data were extracted from electronic medical records, including age, sex, primary and metastatic tumor characteristics, human epidermal growth receptor 2 (HER2), estrogen receptor, progesterone receptor, Ki-67 status, TILs, metastatic sites, immunotherapy regimens, best response to ICIs, disease progression, and survival status.

The CRP levels, absolute neutrophil count, absolute monocyte count, absolute lymphocyte count, and platelet count within the peripheral blood at baseline (within 10 days before treatment with PD-1 inhibitors) and after three courses of PD-1 inhibitors were also recorded. Pretreatment CRP, NLR, LMR, and PLR were calculated before the initiation of PD-1 inhibitor-based therapy and are termed pre-immunotherapy inflammatory biomarkers. Posttreatment CRP (CRP3), NLR (NLR3), LMR (LMR3), and PLR (PLR3) levels after three courses of immunotherapy were calculated between 1 week after the completion of the 3rd course of immunotherapy and the initiation of the next course and termed as postimmunotherapy inflammatory biomarkers. The CRP3 to CRP (CRP3/CRP), NLR3/NLR, and LMR3/LMR ratios were defined as the changes in CRP, NLR, and LMR before and after three courses of immunotherapy.

Statistical analysis

The objective response rate (ORR) was the proportion of patients with confirmed complete response (CR) or partial response (PR) per Response Evaluation Criteria in Solid Tumors 1.1 (RECIST 1.1). PFS was the duration from PD-1 inhibitor initiation to first documented progression or death. OS was the duration from PD-1 inhibitor initiation to last abnormal follow-up or death. The landmark time was defined as “at the completion of 3 course of immunotherapy (approximately 9 weeks).” Inclusion criteria for the landmark analysis were explicitly defined as follows: (1) Completed 3 planned courses of PD-1 inhibitor-based immunotherapy (no dose reduction or treatment interruption exceeding 14 days that impacted treatment completion); (2) No evidence of disease progression (assessed by RECIST v1.1 criteria) during the landmark period; (3) Available complete laboratory data for CRP3, NLR3, LMR3, and PLR3 measured within the landmark window; (4) Available follow-up data.

Optimal cutoff points for CRP, NLR, LMR, and PLR in the survival analysis were obtained using the surv_cutpoint () function of the survminer package in R. Patients were stratified accordingly. Demographic characteristics are expressed as frequencies and percentages for categorical variables and medians and interquartile ranges (IQRs) for quantitative variables. The baseline patient characteristics and best immunotherapy response were compared between groups using the χ² test. Kaplan–Meier method was used for PFS/OS survival analyses, with log-rank test for inflammatory biomarkers and molecular subtypes. Benjamini–Hochberg false discovery rate (FDR) correction was applied to raw log-rank p values to control for multiple testing. FDR-adjusted p < 0.05 was considered statistically significant, and marginal significance (FDR-adjusted p = 0.05–0.10) was retained for exploratory analysis. To address potential immortal-time and guarantee-time bias, a prespecified landmark analysis was performed. We used Little’s missing data were completely random (MCAR) test to assess the missingness pattern, and the result (p > 0.05) confirmed the MCAR. Univariate and multivariate Cox regression identified factors associated with PFS/OS; factors with p < 0.05 in univariate analysis entered multivariate analysis, with HRs and 95% CI calculated. All tests were two-sided (p < 0.05 for significance). Schoenfeld residuals tests verified the proportional hazards assumption (all p > 0.05). Sensitivity analysis excluded early progressors (disease progression/death within 3 immunotherapy courses before landmark time), with repeated Cox regression models to evaluate result robustness. Statistical analyses were performed using SPSS version 25.0 (IBM Corp., Armonk, NY, USA), Prism 6 for Windows Version 6.02 (GraphPad Software, San Diego, CA, USA), and R 4.2.3 (R Foundation for Statistical Computing, Vienna, Austria). The reporting of this study conforms to the STROBE statement.²⁸

Results

Optimal cutoff values for CRP, NLR, LMR, and PLR

Using the surv_cutpoint() function in R, the optimal baseline cutoff values were 15.52 for CRP, 4.79 for NLR, 3.37 for LMR (all p < 0.05), and 115.85 for PLR (p = 0.12). After three cycles of immunotherapy, the optimal cutoffs were 2.95 (CRP3), 1.94 (NLR3), 2.10 (LMR3; all p < 0.05), and 94.44 (PLR3; p = 0.12). For dynamic ratios, the optimal cutoffs were 4.44 for CRP3/CRP and 1.18 for NLR3/NLR (both p < 0.05), whereas LMR3/LMR (0.76, p = 0.19) and PLR3/PLR (1.80, p = 0.12) were not significant. As PLR showed no prognostic relevance, it was excluded from further analyses.

Patient characteristics

The study enrolled 116 patients, including 50 cases of HR+/HER2−, 56 cases of TNBC, 9 cases of HER2-positive (HER2+), and 1 unknown status. The median follow-up was 19.17 months (IQR: 11.50–26.57 months). The median age of patients was 51 years (IQR: 42–58 years). Thirty-nine patients (34%) presented liver metastases, 37 patients (32%) had lung metastases, and 7 (6%) had both lung and liver metastases. Bone was the most common site of metastasis (44, 38%). Fifty-three percent of patients (n = 62) accepted PD-1 inhibitor-containing therapy in the second-line or later settings, and 63 (54%) received taxanes as part of their immunotherapy regimen. A total of 52 patients with available TILs data at initial diagnosis or prior to immunotherapy were stratified into three groups: low (TILs <10%, n = 41), intermediate (TILs 10%—20%, n = 7), and high (TILs ⩾20%, n = 4). The characteristics of the study population are summarized in Table 1.

Table 1.

Patient characteristics.

Basic information	Number	Percentage (%)
Total	116	100
Age	116
Median (range)	51(42–58)
⩽50	57	49
>50	59	51
Molecular subtype	116
HR+/HER2−	50	43
HR−/HER2−	56	48
HER2+	9	8
Unknown	1	1
Clinical stage at initial diagnosis	116
Stage IV	14	12
Stage I–III	102	88
Immunotherapy best response	116
CR/PR	33	28
SD	53	46
Recurrent or metastatic sites	116
Liver	39	34
Lung	37	32
Bone	44	38
Brain	13	11
Number of recurrences or metastases	116
<3	64	55
⩾3	52	45
The number of lines of immunotherapy	116
First-line	54	47
Second line or later	62	53
Immunotherapy combination regimens
Taxanes	63	54
Bevacizumab	25	22
Platinum	31	27
Tumor-infiltrating lymphocytes	52
Low	41	79
Intermediate	7	13
High	4	8

CR, complete response; HER2−, human epidermal growth factor receptor-negative; HER2+, human epidermal growth factor receptor-positive; HR−, hormone receptor-negative; HR+, hormone receptor-positive; PR, partial response; SD, stable disease.

Baseline CRP data were available for 101 of 116 patients, and baseline NLR and LMR data were available for 106 patients. As shown in Table 2, no significant correlations were present between the pre-immunotherapy inflammatory markers (CRP, NLR, and LMR) and clinical characteristics (p > 0.05). The levels of CRP, NLR, and LMR did not significantly differ before and after anti-PD-1 treatment (p > 0.05).

Table 2.

Associations between pre-immunotherapy inflammatory markers (CRP, NLR, LMR) and clinicopathological characteristics.

Characteristics	CRP				NLR				LMR
Characteristics	Number	⩽15.52	>15.52	p-Value	Number	⩽4.79	>4.79	p-Value	Number	⩽3.37	>3.37	p-Value
Median (quartile)	4.27 (1.35–14.95)				2.80 (2.03–4.13)				2.79 (1.97–3.95)
Total	101	78 (77)	23(23)	–	106	85(80)	21(20)	–	106	66(62)	40(38)	–
Age	101				106				106
⩽50	53	43(55)	10(43)	0.33	54	45(53)	9(43)	0.41	54	35(53)	19(48)	0.58
>50	48	35(45)	13(57)		52	40(47)	12(57)		52	31(47)	21(52)
Molecular subtype	100				105				105
HR+/HER2−	43	32(41)	11(50)	0.76	44	33	11	0.60	44	29(44)	15(39)	0.61
HR−/HER2−	49	39(50)	10(45)		53	44	9		53	31(47)	22(56)
HER2+	8	7(9)	1(5)		8	7	1		8	6(9)	2(5)
Clinical stage at initial diagnosis	101				106				106
Stage IV	12	8(10)	4(17)	0.46	13	10(12)	3(14)	0.72	13	10(15)	3(8)	0.36
Stages I–III	89	70(90)	19(83)		93	75(88)	18(86)		93	56(85)	37(92)
Immunotherapy best response	101				106				106
CR/PR	31	28(36)	3(13)	0.05	31	25(30)	6(26)	0.32	31	19(29)	12(30)	0.48
SD	45	33(42)	12(52)		47	38(46)	9(39)		47	27(41)	20(50)
PD	25	17(22)	8(35)		28	20(24)	8(35)		28	20(30)	8(20)
Liver metastasis	101				106				106
No	68	49(63)	19(83)	0.08	69	57(67)	12(57)	0.39	69	39(59)	30(75)	0.10
Yes	33	29(37)	4(17)		37	28(33)	9(43)		37	27(41)	10(25)
Lung metastasis	101				106				106
No	67	51(65)	16(70)	0.71	72	57(68)	15(71)	0.70	72	45(68)	27(68)	0.94
Yes	34	27(35)	7(30)		34	28(32)	6(29)		34	21(32)	13(32)
Bone metastasis	101				106				106
No	63	46(59)	17(74)	0.19	66	53(62)	13(62)	0.10	66	41(62)	25(63)	0.97
Yes	38	32(41)	6(26)		40	32(38)	8(38)		40	25(38)	15(37)
Brain metastasis	101				106				106
No	88	70(90)	18(78)	0.17	94	77(91)	17(81)	0.25	94	58(88)	36(90)	0.99
Yes	13	8(10)	5(22)		12	8(9)	4(9)		12	8(12)	4(40)
Number of recurrences or metastasis	100				103				105
<3	50	41(53)	9(43)	0.33	53	46(54)	7(33)	0.08	53	30(45)	23(58)	0.18
⩾3	50	37(47)	13(57)		52	38(46)	14(67)		52	36(55)	16(42)
Combination regimens including taxanes	101				106				106
No	46	32(41)	14(61)	0.09	48	40(47)	8(38)	0.48	48	32(48)	16(40)	0.40
Yes	55	46(59)	9(39)		58	45(53)	13(62)		58	34(52)	24(60)
Combination regimens including bevacizumab	101				106				106
No	77	59(76)	18(78)	0.80	82	66(78)	16(76)	1.00	82	48(73)	34(85)	0.14
Yes	24	19(24)	5(22)		24	19(22)	5(24)		24	18(27)	6(15)
Combination regimens including platinum	101				106				106
No	75	60(77)	15(65)	0.26	79	66(78)	13(62)	0.14	79	51(77)	28(70)	0.41
Yes	26	18(23)	8(35)		27	19(22)	8(38)		27	15(23)	12(30)
The number of lines of immunotherapy	101				106				106
First-line	45	35(45)	10(43)	0.91	49	42(49)	7(33)	0.19	49	29(44)	20(50)	0.54
Second line or later	56	43(55)	13(57)		57	43(51)	14(67)		57	37(56)	20(50)
Tumor-infiltrating lymphocytes	49			0.16	50			1.00	50			0.25
Low	38	27(71)	11(100)		39	31(76)	8(89)		39	26(81)	13(72)
Intermediate	7	7(18)	0		7	6(14)	1(11)		7	5(16)	2(11)
High	4	4(11)	0		4	4(10)	0		4	1(3)	3(17)

CR, complete response; CRP, C-reactive protein; HER2−, human epidermal growth factor receptor-negative; HER2+, human epidermal growth factor receptor-positive; HR−, hormone receptor-negative; HR+, hormone receptor-positive; LMR, lymphocyte-to-monocyte ratio at baseline; NLR, neutrophil-to-lymphocyte ratio at baseline; PD, progressive disease; PR, partial response; SD, stable disease.

PFS, OS, and ORR of the overall study population

The median PFS (mPFS) was 4.76 months (95% CI: 3.70–5.64 months) and the median OS (mOS) was 20.23 months (95% CI: 15.52–24.94 months). The ORR was 28% (95% CI: 0.20–0.37%) and the disease control rate was 74% (95% CI: 0.66–0.82%).

The prognostic values of different inflammatory biomarkers for PFS, OS, and ORR

Associations between CRP and PFS, OS, and ORR

Kaplan–Meier survival analyses and log-rank tests showed that patients in the low CRP group at baseline had a significantly longer PFS (mPFS: 4.97 vs 3.70 months, p < 0.05) and OS (mOS: 22.03 vs 9.30 months, p < 0.05) than those in the high CRP group (Figure 1(a) and (b)). The ORR was significantly higher in the low CRP group than in the high CRP group (36% vs 13%, p < 0.05).

Figure 1.

Kaplan–Meier curves for PFS and OS. PFS according to baseline CRP (a), baseline NLR (c), baseline LMR (e). OS according to baseline CRP (b), baseline NLR (d), baseline LMR (f).

After three courses of immunotherapy, low CRP3 was significantly associated with longer PFS (mPFS: 8.00 vs 3.27 months, p < 0.05) and OS (mOS: NR vs 12.73 months, p < 0.05; Figure 2(a) and (b)). However, there was no statistically significant difference in ORR (41% vs 27%, p = 0.17).

Figure 2.

Kaplan–Meier curves for PFS and OS. PFS after three courses of treatment with PD-1 inhibitor CRP3 (a), NLR3 (c), LMR3 (e). OS after three courses of treatment with PD-1 inhibitors CRP3 (b), NLR3 (d), LMR3 (f).

Low CRP3/CRP was significantly associated with longer PFS (mPFS: 5.87 vs 2.17 months, p < 0.05) and OS (mOS: 21.73 vs 12.73 months, p < 0.05; Figure 3(a) and (b)). ORR did not significantly differ between the low and high CRP3/CRP groups (34% vs 20%, p = 0.37).

Figure 3.

Kaplan–Meier curves for PFS and OS. PFS associated with changes in CRP, NLR, and LMR levels before and after three courses of treatment with PD-1 inhibitor, CRP3/CRP (a), NLR3/NLR (c), LMR3/LMR (e). OS associated with changes in CRP, NLR, and LMR levels before and after three courses of treatment with PD-1 inhibitor, CRP3/CRP (b), NLR3/NLR (d), LMR3/LMR (f).

Landmark analysis further verified CRP-related indices’ prognostic value. For baseline CRP, the low group trended toward a higher 10-month OS rate (OSR; p = 0.07; Supplementary Figure 1(B)). In contrast, low CRP3 and low CRP3/CRP ratio groups had significantly higher 10-month PFSR and 10-month OSR (both p < 0.05; Supplementary Figure 1(C)–(F)).

Notably, high CRP3 and high CRP3/CRP ratio remained linked to lower survival rates, consistent with primary analysis results.

Associations between NLR and PFS, OS, and ORR

At baseline, the low NLR group had significantly longer PFS (mPFS: 5.60 vs 3.30 months, p < 0.05) and OS (mOS: 21.73 vs 7.51 months, p < 0.05) than those of the high NLR group (Figure 1(c) and (d)). After three courses of immunotherapy, low NLR3 was significantly associated with longer PFS and OS (Figure 2(c) and (d)). A high NLR3/NLR was correlated with shorter PFS (mPFS: 5.22 vs 4.02 months, p < 0.05; Figure 3(c)).

ORR did not significantly differ between the low and high NLR groups at baseline (32% vs 19%, p = 0.25) and after three courses of immunotherapy (36% vs 31%, p = 0.63); likewise, it did not significantly differ between the low and high NLR3/NLR groups (31% vs 34%, p = 0.82).

Landmark analysis further verified NLR-related indices’ prognostic value. For baseline NLR, the low group had a significantly higher 10-month OSR (p = 0.04; Supplementary Figure 2(B)). For NLR3 and NLR3/NLR ratio, the low ratio group had significantly higher 10-month PFSR and OSR (both p < 0.05; Supplementary Figure 2(C)–(F)).

Notably, high baseline NLR, high NLR3, and high NLR3/NLR ratio remained linked to lower survival rates, consistent with the primary analysis results.

Associations between LMR and PFS, OS, and ORR

Before and after three courses of immunotherapy, patients with a high LMR had a significantly prolonged PFS (LMR group: 5.97 vs 4.47 months, p < 0.05; LMR3 group: 5.83 vs 3.19 months, p < 0.05; Figures 1(e) and 2(e)) and OS (LMR group: NR vs 16.33 months, p < 0.05; LMR3 group: 22.03 vs 11.23 months, p < 0.05; Figures 1(f) and 2(f)) compared with those of the low LMR group. The low and high LMR3/LMR groups did not significantly differ in terms of PFS or OS (Figure 3(e) and (f)).

ORR did not significantly differ between the high LMR and low LMR groups before (29% vs 30%, p = 0.89) and after three courses of immunotherapy (32% vs 32%, p = 1.00), nor between low and high LMR3/LMR groups (37% vs 30%, p = 0.63).

Landmark analysis further verified LMR-related indices’ prognostic value. For baseline LMR, the low group trended toward a higher 10-month PFSR (p = 0.08; Supplementary Figure 3(A)) and significantly higher 10-month OSR (p < 0.05; Supplementary Figure 3(B)). For LMR3, the high LMR3 group had significantly higher 10-month PFSR and OSR (p < 0.05; Supplementary Figure 3(C) and 3(D)). Regarding the LMR3/LMR ratio, no significant differences were found in 10-month PFSR or OSR (both p > 0.05; Figure S6(E) and (F)) between the low and high ratio groups. Notably, high LMR and LMR3 remained consistent predictors of higher survival rates, consistent with the primary analysis results.

Associations between TILs and PFS, OS, and ORR

The TILs-high group demonstrated significantly longer PFS compared to the TILs-low and TILs-intermediate groups (mPFS: not reached vs 4.67 months vs 3.90 months; p = 0.04). For OS, none of the four TILs-high patients reached the survival endpoint (data immature; Figure 4). No statistically significant differences in ORR were observed across the TILs-low, -intermediate, and -high groups (p = 0.18).

Figure 4.

Kaplan–Meier survival analysis stratified by TILs levels. (a) PFS. (b) OS. TILs thresholds are defined as: low (<10%), intermediate (10%—20%), and high (⩾20%).

Univariate and multivariate Cox regression analyses of PFS and OS

Univariate Cox regression identified nine significant predictors of PFS: NLR (p = 0.04), LMR (p = 0.03), CRP3 (p < 0.01), NLR3 (p = 0.02), LMR3 (p < 0.01), CRP3/CRP ratio (p < 0.01), NLR3/NLR ratio (p = 0.03), TILs-low (p = 0.06), and TILs-intermediate (p = 0.03).

Multivariate analysis adjusted for clinicopathological covariates revealed three independent predictors of PFS. Low NLR3 (HR = 0.19, 95% CI: 0.05–0.81; p = 0.02) and reduced CRP3/CRP ratio (HR = 0.29, 95% CI: 0.07–0.98; p = 0.05) were significantly associated with longer PFS. TILs-low was associated with shorter PFS (HR = 14.26, 95% CI: 1.38–147.13; p = 0.03; Table 3).

Table 3.

Predictive factors for PFS by univariate and multivariate analyses.

Characteristic	Univariate analyses			Multivariate analyses
Characteristic	HR	95% CI	p-Value	HR	95% CI	p-Value
Age (⩽50 vs >50)	0.80	0.54–1.12	0.28
Molecular subtype			0.77
HR(+)/HER-2(−)	0.77	0.36–1.65	0.51	0.53	0.16–1.69	0.28
HR(−)/HER-2(−)	0.76	0.36–1.61	0.47	0.97	0.32–2.99	0.96
HER-2(+)
Combination regimens including taxanes (yes vs no)	0.71	0.47–1.05	0.09
Combination regimens including platinum (yes vs no)	0.69	0.43–1.09	0.11
Combination regimens including bevacizumab (yes vs no)	1.16	0.72–1.87	0.53
Liver metastasis (yes vs no)	1.32	0.87–1.99	0.19
Lung metastasis (yes vs no)	1.12	0.73–1.69	0.61
Bone metastasis (yes vs no)	1.40	0.93–2.11	0.11
Brain metastasis (yes vs no)	1.00	0.55–1.85	0.99
Number of metastases (<3 vs ⩾3)	0.73	0.49–1.10	0.13
First-line treatment (yes vs no)	0.81	0.54–1.20	0.29
CRP (⩽15.52 vs >15.52)	0.61	0.37–1.00	0.05	1.53	0.42–5.60	0.52
NLR (⩽4.79 vs >4.79)	0.60	0.36–0.98	0.04	2.25	0.69–7.32	0.18
LMR (⩽3.37 vs >3.37)	1.67	1.06–2.62	0.03	0.78	0.29–2.08	0.62
CRP3 (⩽2.95 vs >2.95)	0.40	0.24–0.66	0.00	0.36	0.12–1.08	0.07
NLR3 (⩽1.94 vs >1.94)	0.51	0.29–0.88	0.02	0.19	0.05–0.81	0.02
LMR3 (⩽2.10 vs >2.10)	2.30	1.41–3.75	0.00	0.60	0.17–2.13	0.43
CRP3/CRP (⩽4.44 vs >4.44)	0.25	0.13–0.46	0.00	0.29	0.07–0.98	0.05
NLR3/NLR (⩽1.18 vs >1.18)	0.59	0.37–0.94	0.03	0.98	0.26–3.66	0.97
LMR3/LMR (⩽0.76 vs >0.76)	1.13	0.71–1.80	0.62
Tumor-infiltrating lymphocytes
Low	6.91	0.94–50.72	0.06	14.26	1.38–147.13	0.03
Intermediate	10.39	1.27–87.36	0.03	4.43	0.46–48.83	0.20
High

CR, complete response; CRP, C-reactive protein; CRP3, C-reactive protein after 3 courses of immunotherapy; CRP3/CRP, CRP3/CRP ratio; HER2−, human epidermal growth factor receptor-negative; HER2+, human epidermal growth factor receptor-positive; HR−, hormone receptor-negative; HR+, hormone receptor-positive; LMR, lymphocyte-to-monocyte ratio at baseline; LMR3, lymphocyte-to-monocyte ratio after 3 courses of immunotherapy; LMR3/LMR, LMR3/LMR ratio; NLR, neutrophil-to-lymphocyte ratio at baseline; NLR3, neutrophil-to-lymphocyte ratio after 3 courses of immunotherapy; NLR3/NLR, NLR3/NLR ratio; PFS, progression-free survival.

Univariate Cox regression analyses showed that liver metastasis (p = 0.04), number of metastatic sites (p = 0.01), CRP (p < 0.01), NLR (p < 0.01), LMR (p < 0.01), CRP3 (p < 0.01), NLR3 (p < 0.01), LMR3 (p < 0.01), and CRP3/CRP (p < 0.01) were associated with OS. A multivariate Cox regression analysis indicated that low CRP (HR = 0.13; 95% CI: 0.04–0.38; p < 0.01) and low NLR (HR = 0.34; 95% CI: 0.15–0.79; p = 0.01) at baseline were associated with longer OS. OS was shorter in the high CRP3/CPR group (HR = 0.32, 95% CI: 0.12–0.85, p = 0.02) and high NLR3 group (HR = 0.31, 95% CI: 0.11–0.85, p = 0.02) than in other groups, as shown in Table 4.

Table 4.

Predictive factors for OS by univariate and multivariate analysis.

Characteristics	Univariate analyses			Multivariate analyses
Characteristics	HR	95% CI	p-Value	HR	95% CI	p-Value
Age (⩽50 vs >50)	0.76	0.44–1.32	0.33
Molecular subtype			0.62
HR(+)/HER-2(−)	0.88	0.34–2.31	0.80	0.50	0.16–1.54	0.23
HR(−)/HER-2(−)	0.69	0.26–1.84	0.46	0.41	0.13–1.31	0.13
HER-2(+)
Combination regimens including taxanes (yes vs no)	1.10	0.64–1.90	0.73
Combination regimens including platinum (yes vs no)	1.00	0.54–1.85	0.99
Combination regimens including bevacizumab (yes vs no)	1.46	0.76–2.81	0.25
Liver metastasis (yes vs no)	1.79	1.02–3.16	0.04	1.50	0.72–3.13	0.27
Lung metastasis (yes vs no)	1.16	0.66–2.03	0.60
Bone metastasis (yes vs no)	1.66	0.95–2.88	0.08
Brain metastasis (yes vs no)	2.00	0.97–4.13	0.06
Number of metastases (<3 vs ⩾3)	0.46	0.25–0.83	0.01	0.67	0.32–1.40	0.28
First-line treatment (yes vs no)	1.24	0.72–2.14	0.45
CRP (⩽15.52 vs >15.52)	0.19	0.10–0.35	0.00	0.13	0.04–0.38	0.00
NLR (⩽4.79 vs >4.79)	0.31	0.16–0.59	0.00	0.34	0.15–0.79	0.01
LMR (⩽3.37 vs >3.37)	2.91	1.54–5.49	0.00	1.48	0.64–3.42	0.36
CRP3 (⩽2.95 vs >2.95)	0.30	0.15–0.61	0.00	0.79	0.29–2.16	0.64
NLR3 (⩽1.94 vs >1.94)	0.23	0.09–0.54	0.00	0.31	0.11–0.85	0.02
LMR3 (⩽2.10 vs >2.10)	3.58	1.93–6.65	0.00	1.18	0.54–2.54	0.68
CRP3/CRP (⩽4.44 vs >4.44)	0.36	0.19–0.71	0.00	0.32	0.12–0.85	0.02
NLR3/NLR (⩽1.18 vs >1.18)	0.56	0.31–1.0	0.06
LMR3/LMR (⩽0.76 vs >0.76)	1.37	0.74–2.51	0.32

Associations between molecular subtype and OS, PFS, and ORR

There were no statistically significant differences in PFS, OS, or ORR among patients with different subtypes, as shown in Figure 5. A subgroup analysis of HER2-nonexpressing (HER2 IHC 0) and HER2-low groups revealed that OS in the HER2-low expression group was significantly longer than that in the HER2-nonexpressing group (27.83 vs 16.60 months, p = 0.03; Figure 6).

Figure 5.

Kaplan–Meier curves for PFS and OS for different molecular subtype. (a) PFS. (b) OS.

Figure 6.

Kaplan–Meier curves for PFS and OS for different HER2 expression level. (a) PFS. (b) OS.

Inflammatory biomarkers and TILs demonstrate stratified prognostic capacity

Based on prior findings, we hypothesized that NLR3 and CRP3/CRP, combined with TILs, could improve PFS and OS prediction. Patients were stratified by TILs infiltration levels into intermediate/high infiltration (TILs-rich, TILs ⩾ 10%) and low infiltration (TILs-deficient, TILs < 10%) groups. They were also categorized by NLR3 (⩽1.94: low; >1.94: high) and CRP3/CRP (⩽4.44: low; >4.44: high). Four combined categories of NLR3 or CRP3/CRP with TILs were analyzed for prognostic value.

PFS analysis stratified by NLR3 and TILs status (Figure 7(a)) showed the shortest mPFS in TILs-deficient-NLR3-high group (2.89 months), while the TILs-rich-NLR3-high group had 5.92 months. In contrast, the TILs-deficient-NLR3-low group showed an mPFS of 8.00 months (p = 0.05). Although the mPFS for the TILs-rich-NLR3-low group was not reached, the extremely small sample size (n = 1) rendered interpreting this group’s results impractical. Figure 7(b) illustrates the prognostic impact of CRP3/CRP and TILs on PFS. Patients in the TILs-deficient-CRP3/CRP-high group had the shortest PFS, with an mPFS of only 1.90 months. The TILs-deficient-CRP3/CRP-low group had a relatively improved mPFS of 4.83 months, while in the TILs-rich-CRP3/CRP-low group patients had an mPFS of 8.57 months (p < 0.01).

Figure 7.

Kaplan–Meier analysis of PFS and OS stratified by TILs status combined with NLR3 or CRP3/CRP ratios. (a) PFS based on TILs and NLR3. (b) PFS based on TILs and CRP3/CRP. (c) OS based on TILs and NLR3. (d) OS based on TILs and CRP3/CRP. TILs thresholds are defined as: deficient (<10%) and rich (>10%).

Regarding OS, NLR3-low groups had unreached median survival. The TILs-deficient-NLR3-high group showed an mOS of 11.43 months, and the TILs-rich-NLR3-high group showed a higher mOS of 21.73 months (p = 0.06). Although the difference was not statistically significant, a trend toward improved OS was observed in the TILs-rich-NLR3-high cohort, and the patients of TILs-rich/NLR3-low cohort also exhibited longer OS (Figure 7(c)). CRP3-low group had longer mOS than CRP-high group (p = 0.05; Figure 7(d)).

We concluded that patients with high NLR3 and low TILs, as well as patients with faster-growing CRP and low TILs, had a worse prognosis. But due to limited sample size (52/116 patients with available TILs data, only 4 in the original “high” TILs group), this analysis is underpowered. Consequently, these results are hypothesis-generating only and must be interpreted with extreme caution.

Discussion

To our knowledge, this is the first study to systematically evaluate multiple pre- and postimmunotherapy inflammatory markers in ABC. We demonstrated that baseline CRP, NLR, and TILs are independent predictors of response to PD-1 inhibitors. In addition, we found that integrating dynamic inflammatory indices (specifically the CRP3/CRP ratio and NLR3) with TILs assessment may provide further refinement in prognostic discrimination for survival outcomes in ABC patients undergoing immunotherapy.

Our findings are consistent with prior analyses in other tumor types. For example, in a pooled analysis of three large phase II/III randomized melanoma trials, patients with high baseline CRP had significantly shorter OS compared with those with low CRP.²⁹ Similarly, elevated baseline NLR has been linked to poor prognosis in diverse solid tumors, including lung cancer and melanoma.^17,30–32 In contrast, a small trial of 34 TNBC patients treated with ICIs reported worse PFS in the low NLR group,²³ differing from our results—likely due to limited sample size. Importantly, no previous breast cancer study has simultaneously examined the prognostic value of CRP and NLR across subtypes and outcomes (PFS and OS). Our analysis provides novel evidence that both markers, when measured before and during treatment, can inform prognosis in ABC, and could serve as inexpensive, minimally invasive biomarkers readily obtainable from peripheral blood.

Our exploratory analysis of TILs in the cohort aligns with previous findings,⁷ reinforcing their established role in predicting ICI efficacy in breast cancer. Our study, however, reveals a critical additional layer of prognostic information derived from dynamic systemic inflammatory markers. Regardless of baseline TILs levels, patients with lower NLR3 or CRP3/CRP ratios had better survival, suggesting that dynamic inflammatory changes may complement TILs status to refine patient stratification. This suggests that while TILs provide an essential snapshot of the local tumor immune context, they may not fully capture the evolving systemic host response to therapy. The dynamic changes in inflammatory biomarkers appear to reflect this systemic dimension, offering prognostic value that is complementary to, and independent of, the initial state of the tumor microenvironment. Notably, this combined classifier finding is limited by the small number of patients with available TIL data. While our initial observations regarding TILs-inflammation combined profiles offer tentative insights, their underpowered nature underscores the need for extreme caution in interpretation. Consequently, developing a more nuanced and powerful patient stratification approach by integrating these dynamic systemic markers with baseline TILs assessment will require validation in larger-scale prospective studies.

The clinical relevance of CRP and NLR is supported by their mechanistic links to systemic inflammation. Cancer-associated inflammation influences tumorigenesis, progression, and prognosis.^33,34 CRP and neutrophils are widely recognized mediators of inflammatory responses, and elevated levels typically reflect an augmented systemic inflammatory state. In melanoma, CRP has been shown to suppress proliferation and effector functions of activated CD4⁺ and CD8⁺ T cells^35,36 and to modulate immune and signaling pathways in macrophages and monocytes.^37–39 In addition, the IL-6–STAT3 inflammatory axis is known to contribute to tumor-associated immune suppression.¹³ Because CRP production is largely driven by circulating IL-6, systemic CRP levels may partially reflect this immunosuppressive inflammatory environment.

NLR, one of the most commonly used systemic inflammation indices, has been associated with poorer outcomes in ICI-treated patients across malignancies.^32,40 However, another study involving 34 cases of TNBC treated with ICIs showed that PFS is significantly worse in the low NLR group,²³ contrary to our results. This may be attributed to the small sample size of the previous study. However, neutrophils are functionally heterogeneous, capable of both tumor-promoting and tumor-suppressive effects depending on their phenotypic polarization,⁴¹ which may underlie discrepancies among NLR studies. We did not identify LMR as an independent predictor in multivariate analysis, despite prior reports linking higher LMR to improved immunotherapy outcomes.^25,42 Differences in cohort composition, immune microenvironment complexity, and statistical power may contribute to these inconsistencies. Further studies are required to clarify the prognostic utility of LMR.

Beyond baseline inflammatory status, dynamic changes during treatment may provide further insight into treatment-related immune modulation. Absolute CRP levels may vary substantially among individuals because of differences in baseline inflammatory or tumor burden. By normalizing posttreatment CRP values to baseline levels, the CRP3/CRP ratio captures relative changes in systemic inflammation during therapy rather than static host characteristics. Early declines in CRP may reflect attenuation of IL-6-driven inflammatory signaling and improved immune engagement during PD-1 blockade, whereas persistently elevated CRP levels may indicate ongoing inflammatory resistance mechanisms.¹⁶ Therefore, the CRP3/CRP ratio may serve as a dynamic indicator of systemic inflammatory modulation during PD-1 inhibitor therapy.

From a translational perspective, dynamic peripheral inflammatory markers may provide a practical complement to tumor-based immune biomarkers in patients with ABC undergoing immunotherapy. While biomarkers such as PD-L1 expression and TILs are increasingly used to identify biologically selected patient populations, they primarily reflect the local tumor immune microenvironment.^6,7,43 In contrast, peripheral inflammatory markers such as CRP may capture systemic host responses that evolve during treatment. Because CRP testing is inexpensive and widely available in routine clinical practice, monitoring CRP dynamics during PD-1 inhibitor therapy may help identify patients with suboptimal immune engagement who may benefit from closer monitoring or early reassessment of treatment strategies.

Our subgroup analysis revealed significantly longer OS in the HER2-low population than in the HER2-nonexpressing group among patients treated with PD-1 inhibitors. Previous research has reported higher PD-L1 expression in immune cells within HER2-low tumors,⁴⁴ whereas others have suggested weaker immune responses compared to HER2-negative cases.⁴⁵ These conflicting findings raise the possibility that HER2-low breast cancer constitutes a distinct biological subtype, warranting additional investigation into its sensitivity to immunotherapy.

The optimal cut-off values for CRP, NLR, and LMR vary between studies, influenced by differences in patient populations, treatment settings, and statistical methodology.^18,46,47 In our cohort, we determined optimal thresholds using standardized survival‑based methods and confirmed their prognostic validity in multivariate Cox regression. Nonetheless, larger, multicenter datasets are required to establish robust, generalizable cut-off points.

Our study has limitations, including a relatively small, single-center cohort and potential confounding from concurrent chemotherapy toxicity, granulocyte colony-stimulating factor administration, and bone marrow reserve, which may affect hematological parameters. Moreover, we did not compare inflammatory markers with other established or emerging biomarkers for immunotherapy response in breast cancer, such as circulating tumor DNA or TMB. PD-L1 expression was not assessed, owing to practical constraints of this retrospective study: limited availability of archived pathological specimens, prohibitive testing costs for advanced/later-line patients, and evidence from the multicenter TORCHLIGHT study showing no clear correlation between PD-L1 status and PD-1 inhibitor efficacy in this population,³ leading to nonroutine PD-L1 testing in clinical practice in China. This reflects the authentic real-world scenario, and our findings on inflammatory biomarkers offer a practical, accessible prognostic alternative.

Conclusion

CRP and NLR represent promising and readily accessible biomarkers for predicting the efficacy of immunotherapy in patients with ABC. These findings highlight a simple, cost‑effective approach that could facilitate individualized immunotherapy decision‑making. Prospective studies with larger, independent cohorts are warranted to validate these results and refine their clinical applicability.

Supplemental Material

sj-docx-1-tam-10.1177_17588359261443939 – Supplemental material for Dynamic inflammatory biomarkers as prognostic indicators in advanced breast cancer patients receiving PD-1 inhibitors

Supplemental material, sj-docx-1-tam-10.1177_17588359261443939 for Dynamic inflammatory biomarkers as prognostic indicators in advanced breast cancer patients receiving PD-1 inhibitors by Na Wang, Peijun Yi, Qianyi Lu, Yi Zhong, Cong Xue, Ping Zhang, Shusen Wang, Zhongyu Yuan, Yanxia Shi, Hanmu Chen, Fei Xu and Kuikui Jiang in Therapeutic Advances in Medical Oncology

Supplemental Material

sj-docx-2-tam-10.1177_17588359261443939 – Supplemental material for Dynamic inflammatory biomarkers as prognostic indicators in advanced breast cancer patients receiving PD-1 inhibitors

Supplemental material, sj-docx-2-tam-10.1177_17588359261443939 for Dynamic inflammatory biomarkers as prognostic indicators in advanced breast cancer patients receiving PD-1 inhibitors by Na Wang, Peijun Yi, Qianyi Lu, Yi Zhong, Cong Xue, Ping Zhang, Shusen Wang, Zhongyu Yuan, Yanxia Shi, Hanmu Chen, Fei Xu and Kuikui Jiang in Therapeutic Advances in Medical Oncology

Supplemental Material

sj-docx-3-tam-10.1177_17588359261443939 – Supplemental material for Dynamic inflammatory biomarkers as prognostic indicators in advanced breast cancer patients receiving PD-1 inhibitors

Supplemental material, sj-docx-3-tam-10.1177_17588359261443939 for Dynamic inflammatory biomarkers as prognostic indicators in advanced breast cancer patients receiving PD-1 inhibitors by Na Wang, Peijun Yi, Qianyi Lu, Yi Zhong, Cong Xue, Ping Zhang, Shusen Wang, Zhongyu Yuan, Yanxia Shi, Hanmu Chen, Fei Xu and Kuikui Jiang in Therapeutic Advances in Medical Oncology

Footnotes

Acknowledgements

None.

Declarations

ORCID iDs

Na Wang

Shusen Wang

Fei Xu

Kuikui Jiang

Supplemental material

Supplemental material for this article is available online.

References

Lyons

Dickler

Comen

EE.

Checkpoint inhibitors in the treatment of breast cancer [J]. Curr Oncol Rep 2018; 20(7): 51.

Cortes

Cescon

Rugo

, et al. Pembrolizumab plus chemotherapy versus placebo plus chemotherapy for previously untreated locally recurrent inoperable or metastatic triple-negative breast cancer (KEYNOTE-355): a randomised, placebo-controlled, double-blind, phase 3 clinical trial [J]. Lancet 2020; 396(10265): 1817–1828.

Jiang

Ouyang

Sun

, et al. Toripalimab plus nab-paclitaxel in metastatic or recurrent triple-negative breast cancer: a randomized phase 3 trial [J]. Nat Med 2024; 30(1): 249–256.

Schlam

Corti

Sammons

, et al. Checkpoint inhibition for early-stage hormone receptor-positive breast cancer [J]. Expert Opinion Biol Ther 2024; 24(6): 511–520.

Emens

Esteva

Beresford

, et al. Trastuzumab emtansine plus atezolizumab versus trastuzumab emtansine plus placebo in previously treated, HER2-positive advanced breast cancer (KATE2): a phase 2, multicentre, randomised, double-blind trial [J]. Lancet Oncol 2020; 21(10): 1283–1295.

Jiao

, et al. Biomarkers for predicting efficacy of PD-1/PD-L1 inhibitors [J]. Mol Cancer 2018; 17(1): 129.

Rizzo

Ricci

AD.

Biomarkers for breast cancer immunotherapy: PD-L1, TILs, and beyond [J]. Expert Opin Investig Drugs 2022; 31(6): 549–555.

Brummel

Eerkens

De Bruyn

, et al. Tumour-infiltrating lymphocytes: from prognosis to treatment selection [J]. Br J Cancer 2023; 128(3): 451–458.

Elinav

Nowarski

Thaiss

, et al. Inflammation-induced cancer: crosstalk between tumours, immune cells and microorganisms [J]. Nat Rev Cancer 2013; 13(11): 759–771.

10.

Karin

Nuclear factor-kappaB in cancer development and progression [J]. Nature 2006; 441(7092): 431–436.

11.

Antonangeli

Natalini

Garassino

, et al. Regulation of PD-L1 expression by NF-κB in cancer [J]. Front Immunol 2020; 11: 584626.

12.

Mantovani

Allavena

Sica

, et al. Cancer-related inflammation [J]. Nature 2008; 454(7203): 436–444.

13.

Pardoll

Jove

STATs in cancer inflammation and immunity: a leading role for STAT3 [J]. Nat Rev Cancer 2009; 9(11): 798–809.

14.

Johnson

O’Keefe

Grandis

. Targeting the IL-6/JAK/STAT3 signalling axis in cancer [J]. Nat Rev Clin Oncol 2018; 15(4): 234–248.

15.

Grivennikov

Greten

Karin

Immunity, inflammation, and cancer [J]. Cell 2010; 140(6): 883–899.

16.

Huang

Zhang

Nomograms confirm serum IL-6 and CRP as predictors of immune checkpoint inhibitor efficacy in unresectable hepatocellular carcinoma [J]. Front Immunol 2024; 15: 1329634.

17.

Alessi

Ricciuti

Alden

, et al. Low peripheral blood derived neutrophil-to-lymphocyte ratio (dNLR) is associated with increased tumor T-cell infiltration and favorable outcomes to first-line pembrolizumab in non-small cell lung cancer [J]. J Immunother Cancer 2021, 9(11): e003536.

18.

Valero

Lee

Hoen

, et al. Pretreatment neutrophil-to-lymphocyte ratio and mutational burden as biomarkers of tumor response to immune checkpoint inhibitors [J]. Nat Commun 2021; 12(1): 729.

19.

Gao

Zhang

, et al. Pretreatment neutrophil-to-lymphocyte ratio as a prognostic biomarker in unresectable or metastatic esophageal cancer patients with anti-PD-1 therapy [J]. Front Oncol 2022; 12: 834564.

20.

Gawiński

Michalski

Mróz

, et al. Correlation between lymphocyte-to-monocyte ratio (LMR), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR) and tumor-infiltrating lymphocytes (TILs) in left-sided colorectal cancer patients [J]. Biology 2022; 11(3): 385.

21.

Templeton

Mcnamara

Šeruga

, et al. Prognostic role of neutrophil-to-lymphocyte ratio in solid tumors: a systematic review and meta-analysis [J]. J Natl Cancer Inst 2014; 106(6): dju124.

22.

Pérez-García

Llombart-Cussac

Cortés

, et al. Pembrolizumab plus eribulin in hormone-receptor-positive, HER2-negative, locally recurrent or metastatic breast cancer (KELLY): an open-label, multicentre, single-arm, phase II trial [J]. Eur J Cancer 2021; 148: 382–394.

23.

Han

Wang

Sun

, et al. Genomic landscape and peripheral blood biomarkers of advanced triple-negative breast cancer treated with immune checkpoint blockade: an exploratory analysis of the TQB2450-Ib-07 trial [J]. J Clin Oncol 2022; 40(16_suppl): 1080.

24.

Tolaney

Barroso-Sousa

Keenan

, et al. Effect of eribulin with or without pembrolizumab on progression-free survival for patients with hormone receptor-positive, ERBB2-negative metastatic breast cancer: a randomized clinical trial [J]. JAMA Oncol 2020; 6(10): 1598–1605.

25.

Sekine

Kanda

Goto

, et al. Change in the lymphocyte-to-monocyte ratio is an early surrogate marker of the efficacy of nivolumab monotherapy in advanced non-small-cell lung cancer [J]. Lung Cancer (Amsterdam, Netherlands) 2018; 124: 179–188.

26.

Denkert

Von Minckwitz

Darb-Esfahani

, et al. Tumour-infiltrating lymphocytes and prognosis in different subtypes of breast cancer: a pooled analysis of 3771 patients treated with neoadjuvant therapy [J]. Lancet Oncol 2018; 19(1): 40–50.

27.

Emens

Molinero

Loi

, et al. Atezolizumab and nab-paclitaxel in advanced triple-negative breast cancer: biomarker evaluation of the IMpassion130 study [J]. J Natl Cancer Inst 2021; 113(8): 1005–1016.

28.

Von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies [J]. J Clin Epidemiol 2008; 61(4): 344–349.

29.

Laino

Woods

Vassallo

, et al. Serum interleukin-6 and C-reactive protein are associated with survival in melanoma patients receiving immune checkpoint inhibition [J]. J Immunother Cancer 2020; 8(1): e000842.

30.

Ouyang

Xiao

Huang

, et al. Baseline and early changes in the neutrophil–lymphocyte ratio (NLR) predict survival outcomes in advanced colorectal cancer patients treated with immunotherapy [J]. Int Immunopharmacol 2023; 123: 110703.

31.

Bartlett

Flynn

Panageas

, et al. High neutrophil-to-lymphocyte ratio (NLR) is associated with treatment failure and death in patients who have melanoma treated with PD-1 inhibitor monotherapy [J]. Cancer 2020; 126(1): 76–85.

32.

Guven

Sahin

Erul

, et al. The association between early changes in neutrophil–lymphocyte ratio and survival in patients treated with immunotherapy [J]. J Clin Med 2022; 11(15): 4523.

33.

Gabrilovich

Ostrand-Rosenberg

Bronte

Coordinated regulation of myeloid cells by tumours [J]. Nat Rev Immunol 2012; 12(4): 253–268.

34.

Diakos

Charles

Mcmillan

, et al. Cancer-related inflammation and treatment effectiveness [J]. Lancet Oncol 2014; 15(11): e493–e503.

35.

Yoshida

Ichikawa

Giuroiu

, et al. C reactive protein impairs adaptive immunity in immune cells of patients with melanoma [J]. J Immunother Cancer 2020; 8(1): e000234.

36.

Kött

Zimmermann

Zell

, et al. C-reactive protein flare predicts response to checkpoint inhibitor treatment in melanoma [J]. J Eur Acad Dermatol Venereol 2024; 38(8): 1575–1587.

37.

Fitzgerald

Kagan

JC.

Toll-like receptors and the control of immunity [J]. Cell 2020; 180(6): 1044–1066.

38.

Takeuchi

Akira

Pattern recognition receptors and inflammation [J]. Cell 2010; 140(6): 805–820.

39.

Gong

Liu

Jiang

, et al. DAMP-sensing receptors in sterile inflammation and inflammatory diseases [J]. Nat Rev Immunol 2020; 20(2): 95–112.

40.

Capone

Giannarelli

Mallardo

, et al. Baseline neutrophil-to-lymphocyte ratio (NLR) and derived NLR could predict overall survival in patients with advanced melanoma treated with nivolumab [J]. J Immunother Cancer 2018; 6(1): 74.

41.

Gungabeesoon

Gort-Freitas

Kiss

, et al. A neutrophil response linked to tumor control in immunotherapy [J]. Cell 2023; 186(7): 1448–1464.e20.

42.

Zheng

Xiong

Wang

, et al. Decreased monocyte-to-lymphocyte ratio was associated with satisfied outcomes of first-line PD-1 inhibitors plus chemotherapy in stage IIIB-IV non-small cell lung cancer [J]. Front Immunol 2023; 14: 1094378.

43.

Lopez

Rodas

Villalba-Esparza

Sanmamed

, et al. Biological and clinical significance of tumour-infiltrating lymphocytes in the era of immunotherapy: a multidimensional approach [J]. Nat Rev Clin Oncol 2025; 22(3): 163–181.

44.

Tsang

Tam

, et al. Comprehensive characterization of HER2-low breast cancers: implications in prognosis and treatment [J]. EBioMedicine 2023; 91: 104571.

45.

Van Den Ende

Smid

Timmermans

, et al. HER2-low breast cancer shows a lower immune response compared to HER2-negative cases [J]. Sci Rep 2022; 12(1): 12974.

46.

Chen

Xiao

, et al. Pretreatment neutrophil-to-lymphocyte ratio is correlated with response to neoadjuvant chemotherapy as an independent prognostic indicator in breast cancer patients: a retrospective study [J]. BMC Cancer 2016; 16: 320.

47.

Sacks

Baxter

Campbell

BCV

, et al. Multisociety consensus quality improvement revised consensus statement for endovascular therapy of acute ischemic stroke [J]. Int J Stroke 2018; 13(6): 612–632.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.71 MB

0.88 MB