Clinical characteristics of programmed death-1 inhibitors for older patients with advanced pancreatic cancer

Abstract

BACKGROUND

Pancreatic cancer (PC), a highly malignant gastrointestinal cancer, is generally diagnosed at an advanced stage. However, traditional therapies for advanced PC are limited and often unsuitable for older patients.

AIM

To identify clinical predictors in older patients with advanced PC to facilitate individualized treatment.

METHODS

This was a retrospective clinical analysis involving 99 patients aged ≥ 65 years with advanced PC who received programmed death-1 (PD-1) inhibitors at Ningbo Medical Center Lihuili Hospital from January 2019 to January 2025. Univariate and multivariate analyses were conducted to identify clinical predictors for survival outcomes, utilizing blood levels and other clinical information.

RESULTS

The median progression-free survival (PFS) was 4.6 months [95% confidence interval (CI): 3.700-5.800], and the median overall survival (OS) was 6.5 months (95%CI: 5.700-8.200). Multivariate analysis helped identify meaningful clinical differences in PFS and OS across subgroups, including factors such as Eastern Cooperative Oncology Group performance status, prognostic nutritional index, and triglyceride levels. Univariate analysis showed that factors such as the location of primary PC, carbohydrate antigen 199 levels, systemic immune-inflammation, neutrophil-to-lymphocyte ratio, and the combination therapy comprising PD-1 inhibitors and radiotherapy are of significant clinical relevance to both PFS and OS.

CONCLUSION

The treatment of advanced PC with PD-1 inhibitors presented several potential independent clinical predictive indicators of survival outcomes in older patients. This study highlighted the importance of pre-treatment clinical characteristics and hematological variables for predicting treatment outcomes in older patients with PC.

Key Words: Advanced pancreatic cancer; Older adults; Programmed death-1 inhibitors; Eastern Cooperative Oncology Group performance status; Prognostic nutritional index; Triglyceride

Core Tip: This study identified prognostic factors for older patients with advanced pancreatic cancer treated with programmed death-1 (PD-1) inhibitors. The findings indicated that PD-1 inhibitor therapy was safe in this older population. Key predictors of treatment prognosis included the Eastern Cooperative Oncology Group performance status, prognostic nutritional index, and triglyceride levels.

INTRODUCTION

Pancreatic cancer (PC), with an alarmingly low overall 5-year survival rate of only 12%, imposes a substantial global health burden[1]. PC is the sixth leading cause of cancer-related deaths worldwide[2]. In 2022, 511000 individuals were diagnosed with PC, and 467000 people died due to this disease[2]. Notably, the diagnosis of PC, whether in the early or late stage, is particularly challenging because of its insidious nature and difficulty in its detection[3].

PC predominantly affects older individuals. In the United States, the median age at which most patients are diagnosed with PC is 71 years[4]. The annual risk of death from PC exceeds 90 per 100000 individuals for those aged over 80 years, compared to fewer than 2 per 100000 for those aged 35-39 years[5]. Among all PC-related deaths, 68.6% occur in older patients[5]. Adults aged 65 years and older are considered to be at the highest risk of developing PC, with men demonstrating poorer outcomes than women[4,5]. Consequently, the mortality risk associated with this disease rises sharply with age.

Traditional therapies for advanced PC primarily include surgical resection, cytotoxic systemic chemotherapy, and radiotherapy; however, these modalities typically extend patient survival by a few months only[6,7]. While immunotherapy has emerged as a key therapeutic strategy for malignant tumors, its efficacy in improving survival outcomes for PC remains limited[8]. Combination regimens involving immunotherapy, whether as part of immunotherapy-only combinations or combined with cytotoxic systemic chemotherapy or chemoradiation, have shown promising potential in preclinical research and early-phase clinical studies[9]. However, the benefits of immune checkpoint blockade treatment for patients aged ≥ 65 years with PC remain unclear.

Ferrat et al[10] identified mortality predictors for older patients with cancer using the Comprehensive Geriatric Assessment. These predictors included functional impairment [e.g., limitations in the activities of daily living or low scores in the Eastern Cooperative Oncology Group performance status (ECOG PS)], mobility deficits, a higher burden of severe comorbidities, and malnutrition. The present study aims to identify potential prognostic indicators derived from clinical characteristics in older patients with advanced PC who are undergoing immune checkpoint blockade therapy. Our objective is to enhance both survival outcomes and the overall efficacy of immunotherapy for this patient population.

MATERIALS AND METHODS

Patients

We retrospectively collected data from patients aged 65 years and older with advanced PC who received antineoplastic therapies based on programmed death-1 (PD-1) inhibitors, either as a monotherapy, or in combination with intensity-modulated radiotherapy or chemotherapy, at Ningbo Medical Center Lihuili Hospital, China, between January 2019 and January 2025.

Patients who met the following inclusion criteria were included in the study: First, aged 65 years or older receiving standard-dose PD-1 inhibitor therapy (e.g., pembrolizumab, nivolumab, sintilimab, and tislelizumab); second, pathologically confirmed PC; third, advanced PC, as classified based on the 8^th edition of the UICC’s TNM Classification of Malignant Tumors[11], encompassing both locally advanced and metastatic subtypes; fourth, peripheral blood inspection results obtained no more than seven working days before starting the anti-PD-1 treatment; and fifth, presence of at least one lesion measurable as per the Response Evaluation Criteria in Solid Tumors (RECIST) 1.1[12]. The exclusion criteria were as follows: First, a history of other malignancies; second, prior administration of other immunotherapies before beginning the anti-PD-1 treatment; third, current diagnosis of autoimmune diseases or other immune system disorders; fourth, insufficient clinical data for analysis; and fifth, severe cardiovascular disease.

Clinical data regarding sex, age, ECOG PS, first-line treatment, chemotherapy combination (administered within the peri-immunotherapy window, defined as ± 7 days), radiotherapy combination (administered within the peri-immunotherapy window), and tumor location were obtained. The peripheral blood analysis involved the determination of the hepatitis B virus (HBV) infection status, carbohydrate antigen 199 (CA199) level (U/mL), high-density lipoprotein (HDL) level (mmol/L), low-density lipoprotein (LDL) level (mmol/L), triglyceride (TG) level (mmol/L), total cholesterol (TC) level (mmol/L), neutrophil count (× 10⁹/L), lymphocyte count (× 10⁹/L), platelet count (× 10⁹/L), and albumin (ALB) level (g/L), all assessed within seven working days prior to initiating the anti-PD-1 treatment. The prognostic nutritional index (PNI) was computed using the formula: 10 × ALB (g/L) + 0.005 × absolute lymphocyte count (× 10⁹/L). The systemic immune-inflammation (SII) index was defined using the formula derived from the peripheral blood analysis: Platelet number (× 10⁹/L) × neutrophil number (× 10⁹/L)/Lymphocyte number (× 10⁹/L).

Outcome assessment

Tumor response was assessed every 8-12 weeks via magnetic resonance imaging or computed tomography, in accordance with the RECIST 1.1 criteria[12]. Using RECIST 1.1, progression-free survival (PFS) was calculated as the period from the commencement of anti-PD-1 therapy until disease progression, inability to tolerate adverse reactions, or all-cause mortality. Overall survival (OS) was determined as the interval from the beginning of PD-1 inhibitor administration to the last follow-up date, inability to tolerate adverse reactions or death from any cause. The patients were followed up via telephone until April 2025. Adverse events (AEs) related to PD-1 inhibitors were assessed from the start of anti-PD-1 therapy, based on the National Cancer Institute Common Terminology Criteria for AEs (version 5.0)[13]. The study was approved by the Ethics Committee of Ningbo Medical Center Lihuili Hospital (approval number: KY2025SL060-01) and adhered to the principles outlined in the Declaration of Helsinki. Written informed consent was waived due to the retrospective nature of the study.

Statistical analysis

Clinical information was described using medians, proportions, and ranges. During the entire follow-up period, PFS and OS were estimated using the Kaplan-Meier approach. The Cox proportional-hazards model was utilized to evaluate independent indicators with a significant impact on PFS and OS, and the hazard ratio (HR) was simultaneously calculated. P < 0.05 was considered statistically significant. For the multivariate analysis (MVA), variables that showed a statistically significant correlation in the univariate analysis (UVA) (P < 0.05) were employed. The factors analyzed comprised the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), SII, PNI, TG, TC, LDL, HDL, and CA199. To evaluate the predictive utility of these factors, receiver operating characteristic (ROC) curves were constructed, and optimal cutoff values were determined using the Youden index. All statistical analyses were performed using SPSS software (version 20.0; SPSS, Chicago, IL, United States).

RESULTS

Patient baseline characteristics

Patient characteristics are outlined in Table 1. The median age of the patients was 71.0 years (span: 65.0-84.0 years), and the median number of PD-1 cycles administered was six. This study included 99 patients [men = 60 (60.6%); women = 39 (39.4%)]. Among these 99 patients, 65 (65.7%) had an ECOG PS of 0-1. Additionally, 90 patients (90.9%) were not infected with HBV. Primary PC was diagnosed in 46 patients (46.5%) with head involvement and in 53 patients (53.5%) with other locations. The majority of patients (70.7%) had not received any prior systemic therapy. However, 24 patients underwent a combination treatment comprising immunotherapy and radiotherapy, while 79 patients received a combination therapy encompassing immunotherapy and chemotherapy.

Table 1 Clinical characteristics of patients (n = 99), n (%).

Characteristic	Patients
Median age, years (range)	71.0 (65-84)
Age, years
< 75	70 (70.7)
≥ 75	29 (29.3)
Sex
Male	60 (60.6)
Female	39 (39.4)
ECOG PS
0-1	65 (65.7)
2	34 (34.3)
HBV infection
Positive	9 (9.1)
Negative	90 (90.9)
Primary pancreatic cancer
Head	46 (46.5)
Other	53 (53.5)
Number of lines of prior systemic therapy
0	70 (70.7)
1	29 (29.3)
Baseline serum CA199 levels (U/mL)
< 173.55	43 (43.4)
≥ 173.55	56 (56.6)
SII
< 531.575	49 (49.5)
≥ 531.575	50 (50.5)
NLR
< 2.1	38 (38.4)
≥ 2.1	61 (61.6)
PLR
< 217.91	74 (74.7)
≥ 217.91	25 (25.3)
PNI
< 45.6	54 (54.5)
≥ 45.6	45 (45.5)
HDL (mmol/L)
< 1.135	55 (55.6)
≥ 1.135	44 (44.4)
LDL (mmol/L)
< 2.26	52 (52.5)
≥ 2.26	47 (47.5)
TC (mmol/L)
< 4.045	56 (56.6)
≥ 4.045	43 (43.4)
TG (mmol/L)
< 1.315	55 (55.6)
≥ 1.315	44 (44.4)
With radiotherapy
Yes	24 (24.2)
No	75 (75.8)
With chemotherapy
Yes	79 (79.8)
No	20 (20.2)

ECOG PS: Eastern Cooperative Oncology Group performance status; HBV: Hepatitis B virus; CA199: Carbohydrate antigen 199; SII: Systemic immune-inflammation; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; PNI: Prognostic nutritional index; HDL: High-density lipoprotein; LDL: Low-density lipoprotein; TC: Total cholesterol; TG: Triglyceride.

Optimal cutoff values for blood markers

ROC curve analysis was performed to identify the optimal threshold values for various peripheral blood markers across the entire patient cohort. The optimal cutoff points for CA199, SII, NLR, PLR, PNI, HDL, LDL, TC, and TG were determined to be 173.55, 531.575, 2.1, 217.91, 45.6, 1.135, 2.26, 4.045, and 1.315, respectively. The associated AUC values were 0.742 [95% confidence interval (CI): 0.640-0.844], 0.598 (95%CI: 0.477-0.719), 0.699 (95%CI: 0.588-0.809), 0.515 (95%CI: 0.395-0.635), 0.722 (95%CI: 0.615-0.828), 0.569 (95%CI: 0.439-0.699), 0.509 (95%CI: 0.386-0.632), 0.528 (95%CI: 0.405-0.650), and 0.719 (95%CI: 0.616-0.822), respectively. These cutoff values effectively categorized all participant patients into two distinct groups, which were then utilized to predict differences in PFS and OS.

Patient survival

At the final follow-up, 29 participants were alive, while 70 had died. Based on the survival statistics, the median values of PFS and OS for the participants were determined to be 4.6 months (95%CI: 3.700-5.800) and 6.5 months (95%CI: 5.700-8.200), respectively. Factors such as age, sex, HBV infection, PLR, LDL level, TC level, and combination chemotherapy were not associated with the PFS and OS. However, statistical analysis showed that other factors analyzed in this study were significantly associated with either PFS or OS.

UVA showed that the number of prior systemic therapy lines (P = 0.033; Figure 1A) and HDL levels (P = 0.016; Figure 1B) were significantly linked to PFS (Table 2); however, these factors did not demonstrate a significant correlation with OS (Table 3). In contrast, MVA indicated that neither the number of prior systemic therapy lines nor HDL levels were correlated with PFS.

Figure 1 Kaplan-Meier curves for progression-free survival and overall survival in patients with pancreatic cancer treated with programmed death-1 inhibitors. A: Correlation between the number of prior systemic therapy lines and progression-free survival (PFS); B: Correlation between high-density lipoprotein levels and PFS; C: Correlation between Eastern Cooperative Oncology Group performance status (ECOG PS) and PFS; D: Correlation between ECOG PS and overall survival (OS); E: Correlation between primary tumor location and PFS; F: Correlation between primary tumor location and OS; G: Correlation between carbohydrate antigen (CA199) levels and PFS; H: Correlation between CA199 levels and OS; I: Correlation between systemic immune-inflammation (SII) index and PFS; J: Correlation between SII index and OS; K: Correlation between neutrophil-to-lymphocyte (NLR) ratio and PFS; L: Correlation between NLR ratio and OS; M: Correlation between prognostic nutritional index (PNI) and PFS; N: Correlation between PNI and OS; O: Correlation between triglyceride (TG) levels and PFS; P: Correlation between TG levels and OS; Q: Correlation between radiotherapy and PFS; R: Correlation between radiotherapy and OS. HDL: High-density lipoprotein; PFS: Progression-free survival; ECOG PS: Eastern Cooperative Oncology Group performance status; OS: Overall survival; CA199: Carbohydrate antigen; SII: Systemic immune-inflammation; NLR: Neutrophil-to-lymphocyte; PNI: Prognostic nutritional index; TG: Triglyceride.

Table 2 Univariate and multivariate analyses of progression-free survival among patients with advanced pancreatic carcinoma.

Clinical variable	n	PFS univariate			PFS multivariate
		Median (months)	95%CI	P value	HR	95%CI	P value
Age, < 75 years	70	4.2	2.75-5.65	0.190
Sex, male	60	4.9	2.69-7.11	0.079
ECOG PS				< 0.001	0.268	0.124-0.582	0.001
0-1	65	8.6	6.32-10.88
2	34	1.7	1.24-2.16
HBV infection				0.698
Positive	9	5.1	3.64-6.56
Negative	90	5.5	3.77-7.23
Primary tumor location				0.030	1.380	0.810-2.351	0.236
Pancreatic head	46	4.5	3.02-5.98
Body & tail	53	6.0	3.97-8.03
Number of prior systemic therapy lines				0.033	0.763	0.579-1.006	0.055
0	70	6.3	4.63-7.98
1	29	2.6	1.37-3.83
Baseline serum CA199 levels (U/mL)				< 0.001	0.798	0.458-1.393	0.428
< 173.55	43	8.5	5.63-11.37
≥ 173.55	56	3.3	2.27-4.33
SII				0.001	0.682	0.321-1.452	0.321
< 531.575	49	8.5	6.14-10.86
≥ 531.575	50	3.3	1.84-4.76
NLR				< 0.001	0.714	0.312-1.634	0.425
< 2.1	38	8.6	6.16-11.04
≥ 2.1	61	3.3	2.01-4.59
PLR				0.076
< 217.91	74	6.3	5.07-7.53
≥ 217.91	25	4.0	3.18-4.82
PNI				< 0.001	2.988	1.633-5.469	< 0.001
< 45.6	54	3.3	2.01-4.59
≥ 45.6	45	8.5	6.00-11.00
HDL (mmol/L)				0.016	1.027	0.744-1.417	0.874
< 1.135	55	3.9	2.49-5.31
≥ 1.135	44	7.1	3.83-10.38
LDL (mmol/L)				0.559
< 2.26	52	4.9	2.48-7.32
≥ 2.26	47	5.5	3.29-7.71
TC (mmol/L)				0.472
< 4.045	56	4.5	2.77-6.23
≥ 4.045	43	6.3	4.96-7.64
TG (mmol/L)				< 0.001	0.343	0.167--0.706	0.004
< 1.315	55	8.5	6.08-10.92
≥ 1.315	44	2.1	1.36-2.84
With radiotherapy				0.019	0.941	0.489-1.810	0.855
Yes	24	7.1	4.07-10.13
No	75	4.0	2.79-5.21
With chemotherapy				0.123
Yes	79	6.0	4.80-7.20
No	20	1.9	1.02-2.78

ECOG PS: Eastern Cooperative Oncology Group performance status; HBV: Hepatitis B virus; CA199: Carbohydrate antigen 199; SII: Systemic immune-inflammation; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; PNI: Prognostic nutritional index; HDL: High-density lipoprotein; LDL: Low-density lipoprotein; TC: Total cholesterol; TG: Triglyceride; PFS: Progression-free survival.

Table 3 Univariate and multivariate analyses of overall survival among patients with advanced pancreatic carcinoma.

Clinical variable	n	OS univariate			OS multivariate
		Median (months)	95%CI	P value	HR	95%CI	P value
Age, < 75 years	70	6.8	4.00-9.60	0.088
Sex, male	60	6.8	5.85-7.76	0.190
ECOG PS				< 0.001	0.229	0.115-0.454	< 0.001
0-1	65	11.0	7.43-14.57
2	34	2.6	1.31-3.89
HBV infection				0.241
Positive	9	9.0	2.23-15.77
Negative	90	7.5	5.67-9.33
Primary tumor location				0.018	1.776	0.988-3.191	0.055
Pancreatic head	46	6.8	4.03-9.57
Body & tail	53	8.6	3.89-13.31
Number of prior systemic therapy lines				0.148
0	70	8.6	6.02-11.18
1	29	6.0	3.68-8.32
Baseline serum CA199 levels (U/mL)				0.003	0.990	0.563-1.741	0.972
< 173.55	43	10.6	8.70-12.50
≥ 173.55	56	5.8	4.05-7.55
SII				0.001	0.502	0.237-1.063	0.072
< 531.575	49	10.4	8.69-12.11
≥ 531.575	50	4.5	2.56-6.45
NLR				< 0.001	0.548	0.238-1.264	0.158
< 2.1	38	10.6	5.14-16.07
≥ 2.1	61	5.1	2.90-7.30
PLR				0.170
< 217.91	74	9.0	6.36-11.64
≥ 217.91	25	6.8	5.50-8.10
PNI				< 0.001	2.715	1.476-4.994	0.001
< 45.6	54	5.1	2.87-7.33
≥ 45.6	45	10.4	9.20-11.60
HDL (mmol/L)				0.061
< 1.135	55	6.8	4.19-9.41
≥ 1.135	44	9.3	6.21-12.39
LDL (mmol/L)				0.616
< 2.26	52	8.3	3.93-12.67
≥ 2.26	47	8.2	5.94-10.46
TC (mmol/L)				0.546
< 4.045	56	6.8	2.75-10.85
≥ 4.045	43	9.0	7.13-10.87
TG (mmol/L)				< 0.001	0.138	0.066-0.288	< 0.001
< 1.315	55	11.5	7.36-15.64
≥ 1.315	44	3.3	2.00-4.60
With radiotherapy				0.012	0.777	0.418-1.446	0.426
Yes	24	11.0	6.37-15.63
No	75	6.8	4.88-8.72
With chemotherapy				0.070
Yes	79	9.0	7.16-10.84
No	20	2.6	0.00-5.45

ECOG PS: Eastern Cooperative Oncology Group performance status; HBV: Hepatitis B virus; CA199: Carbohydrate antigen 199; SII: Systemic immune-inflammation; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; PNI: Prognostic nutritional index; HDL: High-density lipoprotein; LDL: Low-density lipoprotein; TC: Total cholesterol; TG: Triglyceride; OS: Overall survival.

Other clinical characteristics were also analyzed by UVA. The results showed that 65 patients with an ECOG PS of 0-1 exhibited significantly better PFS (P < 0.001) (Figure 1C) and OS (P < 0.001) (Figure 1D). In contrast, 46 patients with primary head PC had worse PFS (P = 0.030) (Figure 1E) and OS (P = 0.018) (Figure 1F). Furthermore, clinical information indicated that patients with CA199 levels exceeding 173.55 U/mL had shorter PFS (P < 0.001) (Figure 1G) and OS (P = 0.003) (Figure 1H). Additionally, patients with SII < 531.575 and NLR < 2.1 had potentially better PFS and OS (Figure 1I-L). Conversely, PNI < 45.6 and TG ≥ 1.315 were correlated with inferior PFS and OS in these patients (Figure 1M-P). The combination treatment comprising PD-1 inhibitors and radiotherapy afforded better PFS (P = 0.019) (Figure 1Q) and OS (P = 0.012) (Figure 1R), as shown by the UVA analysis of the study data.

The MVA of PFS demonstrated significant differences based on the clinical data, including the ECOG PS with a P-value of 0.001 (HR 0.268; 95%CI: 0.124-0.582), PNI with a P-value of < 0.001 (HR 2.988; 95%CI: 1.633-5.469), and TG levels with a P value of 0.004 (HR 0.343; 95%CI: 0.167-0.706) (Table 2). The MVA of OS considered ECOG PS with a P-value of < 0.001 (HR 0.229; 95%CI: 0.115-0.454), PNI with a P-value of 0.001 (HR 2.715; 95%CI: 1.476-4.994), and TG levels with a P-value of < 0.001 (HR 0.138; 95%CI: 0.066-0.288) as potential indicators (Table 3).

Toxicity and safety

This study considered the immune-related AEs (irAEs) of all 99 older patients with advanced PC. The majority of participants were able to tolerate the irAEs associated with PD-1 inhibitors, and no fatalities due to these AEs were reported. During the PD-1 inhibitor therapy, only one patient discontinued the treatment because of a serious rash. Two additional patients experienced grade 3 skin disorders, while two others presented with grade 2 rashes. Other irAEs primarily included fatigue (10.1%), thyroid disorder (6.1%), and liver dysfunction (1.0%).

DISCUSSION

Therapy for older patients with advanced PC remains a difficult clinical challenge, particularly in light of the characteristics of previously reported age groups[14]. Although surgical resection, chemotherapy, and radiotherapy are the typical treatment methods for PC, very few older patients can tolerate these modalities due to low rates of surgical resection and limited efficacy of chemotherapy[15]. Therefore, there is an urgent need to develop biologically targeted alternatives. The PD-1 inhibitor therapy is a novel and effective treatment approach for advanced PC and displays good safety characteristics[16-18]. Even though PD-1 inhibitors have shown promise for older patients with advanced PC, they provide substantial benefits to only a few patients. This issue raises an important question: Which biomarker-defined subgroup achieves clinically significant survival gains from the PD-1 inhibitor therapy in older patients with PC? The present study is an attempt to address this question. Our results revealed that an ECOG PS score of 0-1, PNI ≥ 45.6, and TG levels < 1.315 mmol/L can be identified as independent predictors for the PFS and OS. Older patients with these biomarkers may derive clinical benefits from the PD-1 inhibitor therapy.

In the present study, an ECOG PS of less than 2 was found to be an independent prognostic indicator for both PFS and OS in older individuals diagnosed with advanced PC. A comparable retrospective investigation demonstrated that among the patients with advanced PC treated with a combination of PD-1 inhibitors and chemotherapy, the patients with an ECOG PS score of 1 had a notably shorter PFS than those with an ECOG PS score of 0 (5.6 months vs 8.5 months, P = 0.0002). Additionally, regardless of whether the assessment was based on UVA (P = 0.002) or MVA (P = 0.020), patients with an ECOG PS score of 1 showed a lower objective response rate (ORR)[16]. According to Jung et al[19], in a real-world context involving older patients with metastatic PC undergoing chemotherapy, those with an ECOG PS score of 2 exhibited significantly poorer outcomes (HR = 5.67, P < 0.001). A small retrospective analysis of patients with advanced PC receiving a combined treatment regimen of gemcitabine, nab-paclitaxel, and anti-PD-1 antibodies indicated that ECOG PS significantly influenced both the OS (HR = 8.470, P = 0.003) and PFS (HR = 3.159, P = 0.029)[20]. In yet another study, individuals with an ECOG PS score of 0-1 were found to be more suitable for the treatment, whereas those with an ECOG PS of 3-4 were considered more appropriate for supportive care[21]. Compared to other predictors, the ECOG PS is an inexpensive and clinically accessible parameter. It has also emerged as a prominent predictor for OS in older patients with advanced PC undergoing the PD-1 inhibitor therapy.

The PNI functions as a readily assessable indicator of the nutritional status of patients in clinical practice. A retrospective study involving 530 patients diagnosed with PC identified a negative correlation between the PNI and the SII (R = -0.228, P < 0.001). These two indices are likely to hold considerable importance in the clinical management of PC, especially when it comes to the strategic use of both immunotherapy and targeted therapy[22]. Maehira et al[23] reported that changes in the PNI (HR = 1.064, 95%CI: 1.012-1.119, P = 0.016) are linked to the failure to complete the adjuvant chemotherapy in patients with PC. Another study on perioperative adjuvant chemotherapy in patients with PC found that a PNI of < 44.3 exerts a significant effect on OS: The median survival period among patients with such a low PNI was 25.1 months, in contrast to 39.0 months for those with higher PNI (HR = 1.682, P = 0.028). Low PNI also impacted the relapse-free survival (RFS): The median RFS was 13.1 months for patients with PNI < 44.3, compared with 22.8 months for those with high PNI (HR = 1.559, P = 0.033). Hence, it can be suggested that a low PNI may contribute to the inadequate administration of full-dose adjuvant chemotherapy. For PC patients who underwent surgery and received perioperative adjuvant chemotherapy, this inadequate treatment could ultimately result in disease recurrence and an unfavorable prognosis[24]. PNI is also important for patients preparing for surgery. A low preoperative PNI is an independent prognostic risk indicator associated with shorter OS and a higher rate of postoperative AEs[25,26]. Furthermore, among patients with advanced PC receiving chemotherapy, a PNI of < 43 has been recognized as an independent predictor of unfavorable treatment outcomes[27]. Geng et al[28] confirmed that in advanced PC, a high PNI could predict better survival, with a median OS of 290 days compared to that of 190 days (HR = 0.627, 95%CI: 0.45-0.868, P = 0.003). Similarly, our own study found that a PNI score of ≥ 45.6 was linked to improved OS. Moreover, PNI was highlighted as an independent predictor of OS for older patients with advanced PC undergoing PD-1 inhibitor therapy.

TG metabolism is crucial to cancer progression, because it modulates the tumor microenvironment by reshaping the functional properties of immune cells. A previous study indicated that glycolysis under aerobic conditions serves as the primary pathway through which CD8⁺ T cells produce energy to support their proliferation and the synthesis of cytokines[29]. Therefore, it is plausible to hypothesize that tumor-infiltrating CD8⁺ T cells might lose their normal function within the tumor microenvironment characterized by high lipid levels, low glucose concentrations, and hypoxic conditions. This functional impairment could be linked to the uptake of lipids by CD8⁺ T lymphocytes via CD36 molecules, a process that may ultimately damage these immune cells[30]. The high-lipid microenvironment may also influence the functions of other types of immune cells. For instance, when natural killer cells lose their normal function, the immune system’s surveillance against tumors is impaired, ultimately leading to tumor immune escape and a faster progression of the disease[31]. Wang et al[32] also reported a significant correlation between CD8⁺ T cells and TG levels in patients with hepatocellular carcinoma. Based on existing literature, we postulate that the PD-1 signaling pathway reprograms T cell metabolism to enhance fatty acid oxidation. It is plausible that this metabolic shift fosters a TG-rich tumor microenvironment and the accumulation of lipid-laden macrophages, which subsequently suppress CD8⁺ T cell function. Therefore, TG metabolism may constitute a pivotal mechanistic link between PD-1 signaling and CD8⁺ T cell dysfunction. As mentioned above, a notably negative association between CD8 immuno-reactive scores and TG levels has been documented. In a cohort of 90 patients with PC, those with high TG levels had a significantly poorer prognosis than their counterparts with low TG levels[33]. Likewise, a two-center retrospective study involving 260 patients with PC and 172 patients with non-PC tumors identified TG levels exceeding 1.7 mmol/L as a risk factor for PC (P = 0.007)[34]. To explore the association between abnormal lipid metabolism and PC, Qin et al[35] constructed a high-fat mouse model of PC. They observed significant differences in blood lipid levels between patients with PC and individuals without PC. Their findings confirmed that the serum TG levels in patients with PC at the initial diagnosis were significantly higher than those in individuals without PC. Furthermore, these elevated TG levels showed a positive correlation with neuron-specific enolase levels, which in turn promoted the growth and metastasis of PC cells. Therefore, serum TG levels may act as a predictive PC-associated biomarker[35]. In the present study, TG levels of < 1.315 mmol/L were found to be associated with a prolonged median PFS of 8.5 months (P < 0.001) and an OS of 11.5 months (P < 0.001). Thus, TG may act as an independent predictive indicator for advanced PC in older patients who received PD-1 inhibitor treatment. Therefore, it is necessary to validate the significance of TG through larger and more comprehensive studies in the future.

Inflammation, mediated by immune cells and cytokines, triggers pro-inflammatory pathways and may significantly contribute to the development of PC, both locally and systemically[36]. In a retrospective analysis of 68 patients diagnosed with advanced PC who underwent PD-1 inhibitor therapy along with stereotactic body radiation, patients with an NLR of < 3.2 exhibited a median OS of 27.6 months, compared to 15.6 months in those with an NLR of ≥ 3.2 (P = 0.009)[37]. Additionally, an early-phase clinical trial involving patients with advanced PC receiving immune checkpoint inhibitors revealed that the pre-treatment NLR of ≥ 5 (HR = 1.81, 95%CI: 1.06-3.08, P = 0.029) is associated with a shorter survival period[38]. A meta-analysis further established a correlation between high SII and poor survival outcomes (HR = 1.55, 95%CI: 1.34-1.78, P < 0.001)[39]. Concurrently, another study indicated that SII ≥ 566 and NLR ≥ 2 served as significant negative predictive factors in patients with PC treated with immune checkpoint blockade[40]. We found that lower SII and NLR were associated with better PFS and OS. However, MVA failed to detect a statistically significant difference in this association, possibly due to the small sample size of this study.

Limitations

Several limitations associated with the current study needed to be acknowledged. First, the present study was designed as a retrospective one, characterized by certain inherent constraints common to this type of research design. Second, the sample size may introduce selection or statistical biases, potentially influencing the generalizability and robustness of the study findings. Third, the analysis results may exhibit deviations due to the complex status of microsatellite instability, PD-L1 expression, and tumor mutational burden. Fourth, the variability in subsequent therapies could lead to different survival outcomes. However, the findings also suggested that PD-1 inhibitors were likely effective and safe for older individuals. Therefore, it is imperative to carry out more prospective studies on PD-1 inhibitor treatment for older patients with advanced PC.

CONCLUSION

This study identified ECOG PS, PNI, and TG levels as independent clinical predictive indicators of PFS and OS in older patients diagnosed with advanced PC receiving the PD-1 inhibitor therapy. The UVA revealed that the location of primary PC, along with CA199 levels, SII, NLR, and the combination of PD-1 inhibitors with radiotherapy, exhibited distinct survival outcomes. Notably, in our patient cohort, the effectiveness of the PD-1 inhibitor monotherapy regarding PFS and OS was found to be comparable to that of combination therapy comprising PD-1 inhibitors and chemotherapy. Therefore, the PD-1 inhibitor monotherapy can be a viable treatment choice for older patients with advanced PC, particularly those with an ECOG PS of < 2, a PNI of > 45.6, and TG levels of < 1.315.