Abstract
Background
There are no validated biomarkers that can predict the clinical benefit of immune checkpoint blockers against the programmed cell death protein 1 (PD‐1) treatments in hepatocellular carcinoma (HCC). This study aimed to investigate the prognostic value of inflammation‐immunity‐nutrition score (IINS) in patients with HCC treated with anti‐PD‐1 therapy.
Methods
A consecutive series of 101 HCC patients treated with PD‐1 inhibitors in Sichuan Provincial People's Hospital between January 2018 and August 2020 were enrolled in the retrospective study. IINS (0–6) was constructed based on pretreatment high‐sensitivity C‐reactive protein (hsCRP), lymphocyte (LYM), and albumin (ALB). The patients were divided into high and low IINS groups according to IINS values. Prognostic values of each variable were evaluated with univariate and multivariate time‐dependent Cox regression analyses. Survival curves were calculated and compared using the Kaplan–Meier method and log‐rank test. The prognostic performance of IINS was further compared with that of other traditional prognostic indicators by receiver operating characteristic (ROC) curve and the areas under the ROC curve.
Results
Patients with low IINS had longer overall survival (OS) (HR: 4.711, 95% CI: 1.80–12.37, p = .001) and progression‐free survival (HR: 3.411, 95% CI: 1.79–6.51, p < .0001) than those with high IINS. The multivariate analysis identified IINS (HR: 3.746, 95% CI: 1.05–13.38, p = .042) and tumor number (HR: 5.111, 95% CI: 1.075–24.299, p = .04) as independent prognostic factors. According to ROC analysis, IINS (AUC =0.729, 95% CI: 0.597–0.861, p = .002) presented better prognostic performance than other traditional prognostic indicators. The area of the IINS‐CA19‐9 under the ROC curve (AUC) was higher than that of the IINS or CA19‐9 levels for the prediction of OS.
Conclusion
The results suggest that IINS may be an independent prognostic indicator for HCC patients treated with anti‐PD‐1 therapy. IINS‐CA19‐9 classification may be more effective in predicting clinical benefit of anti‐PD‐1 therapy in HCC patients.
Keywords: biomarkers, hepatocellular carcinoma, inflammation‐immunity‐nutrition score, prognosis, programmed cell death protein 1
Biomarkers are urgently required to identify patients with hepatocellular carcinoma (HCC) likely to benefit from anti‐PD‐1 therapy. This study examines the prognostic value of the inflammation‐immunity‐nutrition score (IINS) in patients with hepatocellular carcinoma (HCC) treated with anti‐PD‐1 therapy. High IINS, multiple tumor number, Child–Pugh stage C, BCLC stage D, extrahepatic metastasis, cycles of anti‐PD‐1, combination with target therapy, and high CA19‐9 level are significantly associated with a shorter overall survival. IINS presented good performance for OS in HCC patients treated with anti‐PD‐1 therapy than other traditional indicators. IINS‐CA19‐9 classification may be more effective in predicting clinical benefit of anti‐PD‐1 treatments in HCC patients.
1. INTRODUCTION
Hepatocellular carcinoma (HCC) is the sixth most common cancer and the third leading cause of cancer death worldwide in 2020, with approximately 906,000 new diagnosed cases and 830,000 deaths. 1 Due to its inapparent symptoms and rapid progression, the majority of HCC patients are diagnosed with advanced stage disease, resulting in failing to radical resection and available HCC treatments merely palliative. 2 Within a decade, immunotherapy targeting immune checkpoints for HCC have grown dramatically and changed the treatment paradigm. Immune checkpoint blockers against PD‐1 and cytotoxic T lymphocyte antigen 4 have been approved for HCC in second‐line treatment, with persisting clinical responses and prolonging survival. 3 However, because of the efficacy of PD‐1 inhibitors varying greatly among individuals, as well as the unsustainable cost burden and immune‐mediated toxicities associated with anti‐PD‐1 therapy, practical and reliable prognostic predictors are urgently need to identify HCC patients who are likely to benefit from anti–PD‐1 therapy.
In immunotherapy, inflammation and innate immunity have a vital role in tumorigenesis, angiogenesis, and immunosuppression. Evidence increasingly suggests that host inflammatory response and immune function are associated with cancer progression and patient survival. 4 Several studies have evaluated that inflammation‐based prognostic scores such as neutrophil‐lymphocyte ratio (NLR), platelet‐lymphocyte ratio (PLR), and prognostic nutritional index (PNI) are predictive biomarkers in patients with HCC treated with anti‐PD‐1 therapy, and higher NLR, PLR, and PNI at baseline are associated with shorter OS and treatment failure. 5 , 6
High‐sensitivity C‐reactive protein (hsCRP) is the most sensitive protein for the detection of systemic inflammation and is also correlated with the prognosis in HCC. 7 , 8 , 9 Lymphocyte (LYM) and albumin (ALB), as two essential components of the inflammation‐based prognostic scores, reflect the systemic immune and nutritional status, respectively. ALB is also closely related to the nutrition and metabolism of cancer patients and their prognosis. 10 , 11 Studies have shown that prognostic scores derived from the combination of several of hsCRP, LYM, and ALB were constructed, such as hsCRP/ALB, hsCRP/LYM, PNI, and were superior to individual indicators in terms of predictive ability. 6 , 8 , 9 As a result, Li et al. initiated that inflammation‐immunity‐nutrition score (IINS), which was simply based on hsCRP, LYM, and ALB, might be prognostic predictor for patients with resectable colorectal cancer, and high IINS presented treatment failure and poor prognosis. 12 However, the prognostic role of IINS in HCC patients treated with anti‐PD‐1 therapy has not been evaluated. Therefore, this study evaluated the prognostic value of IINS on survival outcomes in patients with HCC treated with anti‐PD‐1 therapy using real‐world data.
2. MATERIALS AND METHODS
2.1. Study population
This study enrolled 101 patients who was diagnosed as HCC and treated with PD‐1 inhibitors in Sichuan Provincial People's Hospital between January 1, 2018 and August 31, 2020. Patients were included based on the following criteria: diagnosed with HCC confirmed by pathology; no therapy for primary HCC; had confirmed records of receiving PD‐1 inhibitors; aged from 18 to 75 years; had no other malignant tumors, cardiovascular, and cerebrovascular diseases; and had complete medical and follow‐up data. The exclusion criteria were as follows: having history of malignant tumors or concurrent other malignant tumors; having diseases of the hematologic system; having missing information on hsCRP level, serum ALB level, or LYM counts; Child–Pugh grade D; pre‐existing stress response and inflammatory conditions, such as active or chronic infection; lost to follow‐up within 1 months after anti‐PD‐1 therapy. The flow chart of the study design is showed in Figure 1.
FIGURE 1.
Flow chart of the study design. HCC, hepatocellular carcinoma; PD‐1, programmed cell death protein 1; hsCRP, high‐sensitivity C‐reactive protein; ALB, albumin; LYM, lymphocyte; IINS, inflammation‐immunity‐nutrition score
The study protocol was approved by Sichuan Academy of Medical Sciences & Sichuan Provincial People's Hospital ethics committee (NO2021‐447). Informed consent was obtained from all patients and subjects.
2.2. Data collection
Complete blood parameters of all patients were collected on the day of admission. Prior to receiving PD‐1 inhibitors, those patients received different treatments according to their condition in our hospital, including surgery, transcatheter arterial chemoembolization, hepatic infusion chemotherapy, chemotherapy, and tyrosine kinase inhibitors. Demographic, clinical, and image data were also collected. In addition, according to the response evaluation criteria in solid tumors (RECIST 1.1), tumor assessment and responses to anti‐PD‐1 treatment were performed and classified at baseline and then after every two treatment cycles, which was generally after every 6 weeks. The last follow‐up was on August 31, 2021.
2.3. Calculation
According to the association between each indicators and the patients’ OS, the optimal cutoff points of hsCRP, LYM, and ALB using X‐tile software version 3.6.1 (https://medicine.yale.edu/lab/rimm/research/software/) for survival prediction was determined. Based on 2 cutoffs, hsCRP was divided into three groups: score 0: ≤2.8 mg/L, score 1: >2.8 mg/L and ≤34.3 mg/L, and score 2: >34.3 mg/L, whereas the classification of LYM and ALB was as follows: LYM (score 0: >1.18 × 109/L, score 1: >0.91 × 109/L and ≤1.18 × 109/L, and score 2: ≤0.91 × 109/L); ALB (score 0: >37.7 g/L, score 1: >33.5 g/L and ≤37.7 g/L, and score 2: ≤33.5 g/L) (Supplemental Figure 1). Then, the inflammation‐immunity‐nutrition score (IINS) was generated by summing the scores of hsCRP, LYM, and ALB. In our study, since the median of IINS was 3, the patients were divided into two groups: IINS ≦3 was defined as low IINS group, and IINS >3 was high IINS group, respectively.
2.4. Statistical analysis
Overall survival (OS) defined as the interval from the first PD‐1 inhibitor treatment until death or still alive at the last follow‐up. Progression‐free survival (PFS) was calculated from treatment until disease progression (PD), death, or the last follow‐up (censored). Based on RECIST ver.1.1, the objective response rate (ORR) was the number of patients with complete response (CR) or partial response (PR). The disease control rate (DCR) was defined as the proportion of patients with CR, PR, or stable disease (SD).
The clinical characteristics of patients were compared using the Fisher's exact test for categorical variables and the Wilcoxon rank‐sum test for continuous variables. Continuous and categorical variables were expressed as mean [standard deviation (SD)] and proportions (percentages), respectively. Multivariate time‐dependent Cox proportional PFS and OS hazard ratios (HRs) were fitted based on significant univariate factors. Survival analyses were performed using the Kaplan–Meier method and Log rank test. All P values were from two‐sided tests. All data analyses were performed using SPSS version 27.0 (IBM Corp.,) and GraphPad Prism version 9.2.0 (GraphPad,).
3. RESULTS
3.1. Patient characteristics
A total of 101 patients included in the study, 84 (83.2%), were men, and the median (interquartile ranges, IQRs) age of all patients was 55 (49–63) years. Fifty‐nine (58.4%) patients has hepatitis B virus infection, and the median (IQR) HBV‐DNA copies of whom was 50244.83 (0–249.5) IU/ML. Most patients received another antitumor treatment before anti‐PD‐1 therapy, including surgery (36.6%), transcatheter arterial chemoembolization (TACE) (13.6%), hepatic infusion chemotherapy (HAIC) (19.7%), chemotherapy (9.9%), and tyrosine kinase inhibitors (TKIs) (5%). Forty‐eight (45.5%) patients were treated with PD‐1 monotherapy, and 65 (54.5%) patients were treated with PD‐1 inhibitors simultaneously combined with targeted therapy. The most frequently used PD‐1 inhibitor was camrelizumab (68.3%). Category and dosage of PD‐1 inhibitors used in the two groups are shown in Table S1. The patients with high IINS were more likely to have higher Child–Pugh grade and advanced BCLC Stage, higher hsCRP (mean: 45.76 mg/L), and serum CA19‐9 levels (mean: 1277.37 U/ml), but lower ALB (mean: 31.97 g/L) and LYM (mean: 0.87 × 109/L) levels. The baseline characteristics of the patients are listed in Table 1.
TABLE 1.
Demographic and clinical characteristics of the enrolled patients (N = 101)
| Characteristics | Overall (n = 101) | IINS (0–6) | p value | |
|---|---|---|---|---|
| IINS≤3 (n = 66) | IINS>3 (n = 35) | |||
| Gender (male/female) | 84 (83.2)/17 (16.8) | 54 (81.82)/12 (18.18) | 30 (85.71)/5 (14.29) | .618 |
| Age (years) median (IQR) | 55 (49–63) | 55 (48–63) | 57 (52–64) | .872 |
| BMI (kg/m2) median (IQR) | 22.04 (20.5–24) | 22.24 (20.4–24.22) | 21.80 (20.57–23.5) | .327 |
| Etiology (HBV/HCV/Other) | 59 (58.4)/2 (2)/40 (39.6) | 40 (60.61)/1 (1.52)/22 (33.33) | 19 (54.29)/1 (2.86)/ 17 (48.57) | .483 |
| HBV‐DNA copies median (IQR) | 50244.83 (0–249.5) | 25205.78 (0–166) | 106912.13 (0–1090) | .549 |
| Liver cirrhosis (no/yes) | 49 (48.5)/52 (51.5) | 30 (45.45)/36 (54.55) | 19 (54.29)/16 (45.71) | .398 |
| Child–Pugh grade | .014 | |||
| A | 57 (56.44) | 44 (66.67) | 13 (37.14) | |
| B | 40 (39.6) | 21 (31.82) | 20 (54.29) | |
| C | 4 (3.96) | 1 (1.52) | 2 (8.57) | |
| BCLC Stage | .049 | |||
| B | 61 (60.4) | 44 (66.67) | 17 (48.57) | |
| C | 35 (34.65) | 21 (31.82) | 14 (40) | |
| D | 5 (4.95) | 1 (1.52) | 4 (11.43) | |
| Largest tumor size (cm) | 7.23 (3.59) | 7.19 (3.53) | 7.31 (3.76) | .923 |
| Tumor number (Single/ Multiple) | 41 (40.59)/60 (59.41) | 27 (40.91)/39 (59.09) | 14 (40)/21 (60) | .929 |
| Macrovascular invasion (no/yes) | 68 (67.33)/33 (32.67) | 47 (71.21)/19 (28.79) | 21 (60)/14 (40) | .253 |
| Extrahepatic metastasis (no/yes) | 59 (58.42)/42 (41.58) | 41 (40.59)/25 (59.41) | 18 (51.43)/17 (48.57) | .299 |
| ALB (g/L) | 35.8 (32.5–39.1) | 37.85 (3.85) | 31.97 (4.14) | <.0001 |
| AFP (ng/mL) | 7226.29 (25101.55) | 8171.03 (27113.83) | 5444.78 (21041.52) | .687 |
| CA19‐9 (U/mL) (≤/>18.31) | 11 (5.59)/1010.47 (3091.05) | 7.45 (5.18)/114.01(4741.04) | 7.06 (5.14)/10.9 (3242.30) | .019 |
| Lymphocyte (109/L) | 1.3 (1.58) | 1.53 (1.89) | 0.87 (0.43) | <.0001 |
| hsCRP, mg/L | 26.69 (44.41) | 16.57 (31.43) | 45.76 (57.78) | .001 |
| Cycles of anti‐PD−1 | 10.1 (7.25) | 10.95 (7.61) | 8.49 (6.3) | .12 |
| Previous treatment | .122 | |||
| Surgery | 37 (36.63) | 28 (42.42) | 9 (25.71) | |
| TACE | 22 (21.8) | 9 (13.64) | 13 (37.14) | |
| HAIC | 21 (20.8) | 13 (19.70) | 8 (22.86) | |
| Chemotherapy | 10 (9.9) | 4 (6.06) | 6 (17.14) | |
| TKIs | 5 (5.0) | 3 (4.55) | 2 (5.71) | |
| Combined with target therapy (no/yes) | 46 (45.54)/65 (54.46) | 27 (40.91)/49 (59.09) | 19 (54.29)/16 (45.71) | .062 |
| Cancer progression, n (%) (no/yes) | 60 (59.41)/41 (40.59) | 47 (71.21)/19 (28.79) | 13 (37.14)/22 (62.86) | .001 |
| Death, n (%) (no/yes) | 83 (82.18)/18 (17.82) | 59 (89.39)/7 (10.61) | 24 (68.57)/11 (31.43) | .009 |
Abbreviations: AFP, alpha fetoprotein; ALB, albumin; BCLC, Barcelona Clinic Liver Cancer; BMI, body mass index; CA19‐9, carbohydrate antigen 19–9; HAIC, hepatic infusion chemotherapy; HBV, hepatitis B virus; HCV, hepatitis C virus; hsCRP, high‐sensitivity C‐reactive protein; IINS, inflammation‐immunity‐nutrition score; IQRs, interquartile ranges; PD‐1, programmed cell death protein 1; TACE, transarterial chemoembolization; TKIs, tyrosine kinase inhibitors.
3.2. Treatment response and survival
The median follow‐up time was 11.1 m (95% CI: 3–22.0). The median PFS was 7 m (95% CI: 1.3–22.0), and median OS was 10 m (95% CI: 2–22.0). As summarized in Table 2, no patients achieved complete response, 31 patients (30.7%) achieved partial response, 26 patients (25.7%) had SD, and 42 patients (41.6%) had PD. ORR was 30.7%, and DCR was 56.4%. Low IINS group achieved higher ORR (low vs high IINS, 83.9% vs 16.1%; p = .019), and DCR (low vs high IINS, 48.5% vs 7.9%; p < .001) from anti‐PD‐1 treatment compared with high IINS group. At the end of follow‐up, 41 (40.59%) patients presented cancer progression, and 18 (17.82%) patients died. Moreover, high IINS group presented increasing cancer progression (p = .001) and death risk (p = .009) compared with low IINS group (Table 1).
TABLE 2.
Relationship between IINS groups and response to anti‐PD‐1 treatment
| Best Overall Response | No. of Patients (%) | |||
|---|---|---|---|---|
| Overall n = 101 | IINS≤3 (n = 66) | IINS>3 (n = 35) | p value | |
| CR | 0 (0) | 0 (0) | 0 (0) | |
| PR | 31 (30.7) | 26 (83.9) | 5 (16.1) | .017 |
| SD | 26 (25.7) | 23 (88.5) | 3 (11.5) | .008 |
| PD | 42 (41.6) | 16 (38.1) | 26 (61.9) | <.001 |
| Objective response | 31 (30.7) | 26 (83.9) | 5 (16.1) | .017 |
| Disease control rate | 57 (56.4) | 49 (48.5) | 8 (7.9) | <.001 |
Abbreviations: CR, complete response; PD, disease progression; PR, partial response; SD, stable disease.
3.3. The association between Inflammation‐Immunity‐Nutrition Score and OS
In univariate analysis, IINS (p = .001), tumor number (p = .037), Child–Pugh stage C (p = .041), BCLC stage D (p = .007), extrahepatic metastasis (p = .041), cycles of anti‐PD‐1 (p = .001), and combination with target therapy (p = .005) were all significantly associated with OS. Multivariate analysis revealed that high IINS was associated with significant worse OS, with the multi‐variable‐adjusted hazard ratio (HR) (95% CI) of 3.746 (1.049–13.379, p = .042), and tumor number (p = .04) were also independent prognostic factors (Table 3). There was no statistical significance between different category of PD‐1 inhibitors by univariate analysis (Table S2). The complete information from univariate analysis is listed in Table S3. High IINS was associated with significantly worse OS compared with low IINS (HR 4.711, 95% CI, 1.8–12.4, p = .001) (Figure 2A). In the high IINS and low IINS groups, the median survival was 16.3 months (95% CI, 1.24–22.8) and not reached (95% CI, 2–21.65), respectively.
TABLE 3.
Univariate and multivariate time‐dependent Cox regression analyses of the prognostic factors for OS
| OS | Univariate Analysis | Multivariate Analysis | ||
|---|---|---|---|---|
| Characteristics | HR (95% CI) | p value | HR (95% CI) | p value |
| Child–Pugh grade | ||||
| A | 1 [Ref.] | |||
| B | 1.029 (0.382–2.768) | .955 | 0.744 (0.210–2.630) | .646 |
| C | 5.051 (1.069–23.874) | .041 | 1.090 (0.036–33.223) | .961 |
| BCLC Stage | ||||
| B | 1 [Ref.] | |||
| C | 0.543 (0.153–1.931) | .345 | 0.529 (0.135–2.084) | .363 |
| D | 6.326 (1.664–24.051) | .007 | 2.861 (0.444–18.446) | .269 |
| Tumor number (Single/Multiple) | 3.755 (1.086–12.988) | .037 | 5.111 (1.075–24.299) | .04 |
| Extrahepatic metastasis (no/yes) | 0.609 (0.379–0.979) | .041 | 1.093 (0.318–3.760) | .888 |
| CA19‐9 (U/ml) (≤18.31:>18.31) | 5.808 (1.911–17.652) | .002 | 2.546 (0.726–8.926) | .144 |
| Cycles of anti‐PD−1 | 0.863 (0.788–0.945) | .001 | 0.921 (0.827–1.025) | .131 |
| Combined with target therapy (no/yes) | 0.256 (0.098–0.666) | .005 | 0.344 (0.104–1.131) | .079 |
| IINS | ||||
| High group (IINS≦3) | 1 [Ref.] | |||
| Low group (IINS>3) | 5.858 (2.077–16.519) | .001 | 3.746 (1.049–13.379 | .042 |
Abbreviations: BCLC, Barcelona Clinic Liver Cancer; CA19‐9, carbohydrate antigen 19–9; HR, hazard ratio, CI confidence interval; IINS, inflammation‐immunity‐nutrition score; OS, overall survival; PD‐1, programmed cell death protein 1.
FIGURE 2.
Kaplan–Meier curves of overall survival (A) and progression‐free survival (B) according to inflammation‐immunity‐nutrition Score (IINS) groups
3.4. The association between Inflammation‐Immunity‐Nutrition Score and PFS
In univariate analysis, IINS (p < .0001), BCLC Stage D (p = .019), extrahepatic metastasis (p = .007), serum ALB (≥35:<35) (p = .031), CA19‐9 levels (p = .002), cycles of anti‐PD‐1 (p = .001), and combination with targeted therapy (p = .021) were significantly associated with PFS. In multivariate analysis, high IINS emerged as the powerful unfavorable prognostic factor of PFS (HR: 3.850; 95% CI: 1.007–14.727; p = .049) (Table 4). The complete information from univariate analysis is listed in Table S4. Low IINS was associated with a significantly longer PFS compared with high IINS (HR: 3.411; 95% CI: 1.787–6.512, p < .0001) (Figure 2B). The median PFS was 18.8 (95% CI, 3–23.95) months and 11.5 (95% CI, 4–18.6) months in the low and high IINS groups, respectively.
TABLE 4.
Univariate and multivariate time‐dependent Cox regression analyses of the prognostic factors for PFS
| PFS | Univariate Analysis | Multivariate Analysis | ||
|---|---|---|---|---|
| Characteristics | HR (95% CI) | p value | HR (95% CI) | p value |
| BCLC Stage (B/C/D) | ||||
| B | 1 [Ref.] | |||
| C | 0.861 (0.427–1.735) | .675 | 0.708 (0.173–2.895) | .630 |
| D | 3.671 (1.239–10.873) | .019 | 2.358 (0.460–12.099) | .304 |
| Extrahepatic metastasis (no/yes) | 2.304 (1.261–4.209) | .007 | 1.481 (0.518–4.235) | .464 |
| ALB (g/L) (≦35:<35) | 0.514 (0.280–0.942) | .031 | 1.458 (0.401–5.308) | .567 |
| CA199 (U/ml) (≤18.31:>18.31) | 5.808 (1.911–17.652) | .002 | 2.470 (0.723–8.445) | .149 |
| Cycles of anti‐PD−1 | 0.918 (0.871–0.967) | .001 | 0.920 (0.828–1.023) | .122 |
| Combined with target therapy (no/yes) | 0.496 (0.270–0.913) | .024 | 0.329 (0.108–1.007) | .051 |
| IINS | ||||
| Low group (IINS≦3) |
1 [Ref.] |
|||
| High group (IINS>3) | 3.909 (2.108–7.246) | <.0001 | 3.850 (1.007–14.727) | .049 |
Abbreviations: PFS, progression‐free survival; HR, hazard ratio, CI confidence interval; BCLC, Barcelona Clinic Liver Cancer; ALB, albumin; CA19‐9, carbohydrate antigen 19–9; PD‐1, programmed cell death protein 1; IINS, inflammation‐immunity‐nutrition score.
3.5. Relationships between IINS‐CA19‐9 classification and the prognosis of patients
In our study, similar to IINS groups, with the cutoff values of carbohydrate antigen 19–9 (CA19‐9) (cutoff =18.31), we found patients in the low CA19‐9 group (CA19‐9≦18.31) had a longer OS compared with patients in the high CA19‐9 group (CA19‐9>18.31) (HR:5.799, 95% CI:2.235–15.05; p < .0001, Figure 3A), and longer PFS (HR:1.874, 95% CI:1.023–3.435; p < .05, Figure 3B), respectively. Then, according to IINS‐CA19‐9 classification, patients were divided into different groups as follows: patients with low IINS and low CA19‐9 were group I, patients with high IINS and low CA19‐9 or with low IINS and high CA19‐9 were group II, and patients with high IINS and high CA19‐9 were group III. Group I had a longer OS (p < .0001, Figure 4A) and PFS (p < .0001, Figure 4B) compared with group II and III.
FIGURE 3.
Kaplan–Meier curves were generated to analyze the overall survival (A) and progression‐free survival (B) differences among 101 HCC patients treated with PD‐1 inhibitors who were divided according to the cutoff value of the CA19‐9
FIGURE 4.
Kaplan–Meier survival curves for overall survival (A) and progression‐free survival (B) according to IINS‐CA19‐9 classification
3.6. Comparing the prognostic values of inflammation‐immunity‐nutrition score with other traditional indicators
To compare the prognostic predictive performance of IINS with other traditional indicators, including Child–Pugh grade, α‐fetoprotein (AFP), NLR, PLR, hsCRP/ALB, hsCRP/LYM, and PNI, the AUCs (95% CIs) of the indicators were calculated. ROC analysis was used to further evaluate the effect of these indicators on prognosis in our research. Our results showed that IINS presented better performance for OS in HCC patients treated with anti‐PD‐1 therapy than other traditional indicators (AUC =0.729, 95% CI: 0.597–0.861, specificity =0.722, sensitivity =0.735, p = .002; Table 5, Figure 5A). We further evaluate the effect of IINS‐CA19‐9 classification on prognosis in our research. The ROC analysis showed that the pretreatment IINS‐CA19‐9 scores were more predictive of OS than IINS or serum CA19‐9 levels alone (AUC =0.764, 95% CI: 0.631–0.897, p = .001; Figure 5B).
TABLE 5.
Comparison of the prognostic performance between the indices in overall survival of HCC patients treated with PD‐1 inhibitors
| Variables | Calculations | Cut off value | AUC (95% CI) | Specificity | Sensitivity | p value |
|---|---|---|---|---|---|---|
| IINS | 0.729 (0.597–0.861) | 0.722 | 0.735 | .002 | ||
| CA19‐9 | 18.31 | 0.736 (0.608–0.863) | 0.693 | 0.778 | .002 | |
| IINS‐CA19‐9 | 0.764 (0.631–0.897) | 0.872 | 0.533 | .001 | ||
| Child–Pugh grade | 0.545 (0.395–0.694) | 0.578 | 0.500 | .552 | ||
| AFP | 0.755 | 0.522 (0.370–0.673) | 0.361 | 0.75 | .773 | |
| NLR | Neutrophil count: lymphocyte count | 2.79 | 0.632 (0.502–0.760) | 0.373 | 0.889 | .082 |
| PLR | Platelet count: lymphocyte count | 197.3 | 0.616 (0.463–0.770) | 0.843 | 0.389 | .124 |
| hsCRP/ALB | hsCRP: ALB | 0.7 | 0.562 (0.411–0.713) | 0.735 | 0.389 | .412 |
| hsCRP/LYM | hsCRP: LYM | 2.59 | 0.585 (0.448–0.723) | 0.337 | 0.833 | .258 |
| PNI | ALB +LYM count ×5 | 50.63 | 0.532 (0.380–0.684) | 0.952 | 0.111 | .677 |
Abbreviations: IINS, inflammation‐immunity‐nutrition score; CA19‐9, carbohydrate antigen 19–9; AFP, α‐fetoprotein; NLR, neutrophil‐lymphocyte ratio; PLR, platelet‐lymphocyte ratio; hsCRP, high‐sensitivity C‐reactive protein; LYM, lymphocyte; ALB, albumin; PNI, prognostic nutritional index.
FIGURE 5.
Comparison of ROC curves for the outcome prediction between IINS and other traditional indicators (A), CA19‐9 and IINS‐CA19‐9 scores (B) in HCC patients treated with anti‐PD‐1 therapy
4. DISCUSSION
Programmed cell death protein 1 (PD‐1) inhibitors have emerged as an effective therapeutic approach for hepatocellular carcinoma (HCC), associating with a curable potential and a durable survival due to a substantial heterogeneity, and only a small proportion of HCC patients could benefit from anti‐PD‐1 therapy, resulting in a low objective response rate (ORR) of 10%–25%. 13 Tumor mutational burden, microsatellite instability, PD‐L1 expression level, the number of tumor‐infiltrating lymphocytes, and gene expression characteristics have been evaluated as effective measures to predict treatment response and improve cancer patients’ prognosis, including melanoma, colon cancer, and non‐small cell lung cancer, while the improvement of HCC patients’ prognosis has still remained a main clinical challenge. 13 , 14 , 15 In addition, because of the unavailability of tumor tissues and complex molecular or microscopic analyses, these possible biomarkers have a limited predictive accuracy, and are not practically utilized in clinic. 16 Therefore, it is vital to explore practical and reliable biomarkers that can predict treatment outcomes.
A simple and novel serum biomarker, inflammation‐immunity‐nutrition score (IINS), which is based on hsCRP, LYM, and ALB, was proposed by Li et al. and reported to play strong predictive roles in prognostic outcome for patients with resectable colorectal cancer (CRC). A high IINS was associated with the worse survival of CRC. 12 Furthermore, IINS might serve as an ideal biomarker because of being easily accessible, noninvasive, and cost‐effective.
Similarly, recent studies have suggested that pretreatment neutrophil‐to‐lymphocyte ratio, mutational burden, lung immune prognostic index (LIPI), and inflammation‐based prognostic scores were associated with clinical outcomes and clinical benefits in HCC patients treated with PD‐1 inhibitors. 5 , 15 , 17 , 18 The latest research also suggests that in terms of predictive ability, the PNI score is a discriminatory prognostic indicator for OS in HCC patients with anti‐PD‐1 therapy and was superior to the other inflammation‐based prognostic scores, including Glasgow Prognostic Score (GPS), systemic immune inflammation index (SII), modified Glasgow Prognostic Score (mGPS), prognostic index (PI), CRP‐to‐albumin ratio (CAR), lymphocyte‐to‐monocyte ratio (LMR), and lymphocyte‐to‐CRP ratio (LCR). 6 Different from IINS, most of prognostic biomarkers are different combination of two indexes from serum testing at present, which cannot reflect the immune and nutritive function of the body, in resulting inevitable biases and prediction inaccuracy. 19
To our knowledge, this study is the first to suggest that pretreatment IINS might serve as a robust prognostic score in the treatment of HCC patients with PD‐1 inhibitors. Our results demonstrated that OS and PFS significantly improved in HCC patients who were treated with PD‐1 inhibitors with a low IINS compared with those with a high IINS. IINS‐CA19‐9 classification may be highly effective for predicting the prognosis of HCC patients who were treated with PD‐1 inhibitors. In addition, further comparison revealed that apart from Child–Pugh grade, AFP, NLR, and PLR, hsCRP/ALB, hsCRP/LYM, and PNI were also inferior to IINS in prognostic performance. Thus, IINS could be acknowledged as have better predictive ability than other inflammation‐based prognostic scores.
Tumor‐promoting inflammation is considered as one of the characteristics of cancer development, 20 several studies have demonstrated that inflammatory response is correlated with the efficacy of anti‐PD‐1 therapy in advanced types of cancer, including HCC. 21 , 22 Emerging evidence has shown that inflammation‐based prognostic scores exhibited a promising discriminatory ability in predicting prognosis of HCC patients treated with PD‐1 inhibitors. 5 , 6 HsCRP, LYM, and ALB are the major components of inflammation‐based prognostic scores and were found to be closely correlated with inflammation, immunity, and nutrition, respectively. 23 , 24 Previous studies have also reported the prognostic values of hsCRP/LYM, hsCRP/ALB, and PNI for different types of cancer. 8 , 9 , 24 , 25 The underlying mechanism has not been well clarified. In our study, one possible explanation could be that interleukin (IL)‐6, one of the main inducers of CRP production has been showed to promote tumor growth and metastasis. 26 , 27 , 28 In addition, in the tumor microenvironment, the depletion of lymphocytes, such as CD4‐ and CD8‐positive T cells, would lead to disabling immune surveillance and killing. 29 , 30 A retrospective study indicated that nutrition and metabolism in patients with advanced HCC were closely associated with the efficiency of anti‐PD‐1 treatment and survival benefits, and ALB was positively correlated with efficacy. 31 Low serum ALB levels reflect a state of malnutrition, which would weaken cellular and humoral immunity, phagocytic activities, and other defensive systems in cancer patients. 32
CA19‐9 is a widely used biomarker for pancreatic, biliary, gastric, esophageal, and colonic cancers. 33 Some studies have shown that an elevated CA19‐9 level was associated with poor prognosis of HCC patients who underwent resection or hepatic transplantation. 34 , 35 Furthermore, a prospective study reported that CA19‐9 served as an independent predictor for HCC patients’ survival. 36 CA19‐9 level mainly reflects the pathological conditions, while IINS indicates a patient's overall status, including inflammation, immune, and nutritional status.
In clinical practice, this study suggested that inflammation, immune, and nutritional status in patients with HCC were closely related to anti‐PD‐1 immunotherapy efficacy and survival benefit, clearly imply the presence of a chance for enhanced clinical outcomes or even cure for those HCC patients presenting with lower systemic inflammation loads and favorable immunonutritional status. Hence, reducing levels of systemic inflammation, improving immunity, and nutrition support may be the feasible strategies to enhance treatment efficacy and survival in those patients with high IINS. Furthermore, based on the combined application of IINS and CA19‐9, individualized prediction of prognosis was performed, in which a good response to anti‐PD‐1 therapy could be achieved in group I, followed by moderate and poor responses in groups II and III, respectively. For patients with elevated values of IINS and CA19‐9 (group III), considering the reduced prognostic benefit, it is essential to indicate the possibility of combination with targeted therapy or an early change in therapy. For patients without elevated values of IINS or CA19‐9 (group I), the surgeon should make patients with unresectable HCC achieve down‐staging or radical resection, thereby improving patients’ survival as long as possible.
This study also has certain limitations. First, our research uses a retrospective design. Second, it is a single‐center study with a relatively short median follow‐up (11.1 months). Third, patients received non‐single PD‐1 inhibitors during the treatment, which inevitably caused bias. In addition, the sample size we studied is relatively small. Further multicenter, larger‐scale prospective research is needed to validate our findings.
5. CONCLUSION
Our results suggest that inflammation‐immunity‐nutrition score (IINS) may present as an independent prognostic factor for patients with HCC treated with anti‐PD‐1 therapy. Specifically, pretreatment IINS‐CA19‐9 scores are better than the IINS alone in predicting the prognosis of HCC patients treated with anti‐PD‐1 therapy. A low IINS‐CA19‐9 score (group I, good) suggests that maintenance therapy and close follow‐up should be considered. Identifying patients who may have poor short‐term outcomes can help optimize their treatment strategies, such as combining PD‐1 inhibitors with target therapy or other treatments.
CONFLICT OF INTEREST
The authors declare that there is no conflict of interest.
AUTHOR CONTRIBUTIONS
Z Zhang was responsible for study conception and design, data acquisition, data analysis, and drafting and revision of the manuscript. X Huang was responsible for study conception and design, data analysis and drafting, and revision of the manuscript. D Zhong, Z Dai, J Shang, Y Yao, T Feng, H Zou, and C Lai were responsible for data acquisition. Y Liang was responsible for data analysis and drafting and revision of the manuscript.
ETHICAL APPROVAL
This study was approved by the Sichuan Academy of Medical Sciences & Sichuan Provincial People's Hospital ethics committee (NO2021‐447). Informed consent was obtained from all patients and subjects.
Supporting information
Supplementary Material
ACKNOWLEDGMENT
The authors appreciate the time and effort of all participants.
Zhang Z, Liang Y, Zhong D, et al. Prognostic value of inflammation‐immunity‐nutrition score in patients with hepatocellular carcinoma treated with anti‐PD‐1 therapy. J Clin Lab Anal. 2022;36:e24336. doi: 10.1002/jcla.24336
DATA AVAIL ABILIT Y STATEMENT
Data sharing is not applicable to this article as no datasets were generated or analyzed during the study.
REFERENCES
- 1. Sung H, Ferlay J, Siegel RL, et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71(3):209‐249. [DOI] [PubMed] [Google Scholar]
- 2. Yang JD, Hainaut P, Gores GJ, Amadou A, Plymoth A, Roberts LR. A global view of hepatocellular carcinoma: trends, risk, prevention and management. Nat Rev Gastroenterol Hepatol. 2019;16(10):589‐604. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Macek Jilkova Z, Aspord C, Decaens T. Predictive factors for response to PD‐1/PD‐L1 checkpoint inhibition in the field of hepatocellular carcinoma: current status and challenges. Cancers (Basel). 2019;11(10):1554. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Crusz SM, Balkwill FR. Inflammation and cancer: advances and new agents. Nat Rev Clin Oncol. 2015;12(10):584‐596. [DOI] [PubMed] [Google Scholar]
- 5. Dharmapuri S, Ozbek U, Lin JY, et al. Predictive value of neutrophil to lymphocyte ratio and platelet to lymphocyte ratio in advanced hepatocellular carcinoma patients treated with anti‐PD‐1 therapy. Cancer Med. 2020;9(14):4962‐4970. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Mei J, Sun XQ, Lin WP, et al. Comparison of the prognostic value of inflammation‐based scores in patients with hepatocellular carcinoma after anti‐PD‐1 therapy. J Inflamm Res. 2021;14:3879‐3890. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Ko YJ, Kwon YM, Kim KH, et al. High‐sensitivity C‐reactive protein levels and cancer mortality. Cancer Epidemiol Biomarkers Prev. 2012;21(11):2076‐2086. [DOI] [PubMed] [Google Scholar]
- 8. Liao M, Chen P, Liao Y, et al. Preoperative high‐sensitivity C‐reactive protein to lymphocyte ratio index plays a vital role in the prognosis of hepatocellular carcinoma after surgical resection. Onco Targets Ther. 2018;11:5591‐5600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Oh TK, Choi YR, Cho JY, et al. The high‐sensitivity C‐reactive protein/albumin ratio predicts long‐term oncologic outcomes after curative resection for hepatocellular carcinoma. J Clin Med. 2018;7(6):139. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Konigsbrugge O, Posch F, Riedl J, et al. Association between decreased serum albumin with risk of venous thromboembolism and mortality in cancer patients. Oncologist. 2016;21(2):252‐257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Seebacher V, Grimm C, Reinthaller A, et al. The value of serum albumin as a novel independent marker for prognosis in patients with endometrial cancer. Eur J Obstet Gynecol Reprod Biol. 2013;171(1):101‐106. [DOI] [PubMed] [Google Scholar]
- 12. Li XY, Yao S, He YT, et al. Inflammation‐immunity‐nutrition score: a novel prognostic score for patients with resectable colorectal cancer. J Inflamm Res. 2021;14:4577‐4588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Cheng AL, Hsu C, Chan SL, Choo SP, Kudo M. Challenges of combination therapy with immune checkpoint inhibitors for hepatocellular carcinoma. J Hepatol. 2020;72(2):307‐319. [DOI] [PubMed] [Google Scholar]
- 14. Ready N, Hellmann MD, Awad MM, et al. First‐line nivolumab plus ipilimumab in advanced non–small‐cell lung cancer (checkmate 568): outcomes by programmed death ligand 1 and tumor mutational burden as. Biomarkers. 2019;37(12):992‐1000. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Samstein RM, Lee CH, Shoushtari AN, et al. Tumor mutational load predicts survival after immunotherapy across multiple cancer types. Nat Genet. 2019;51(2):202‐206. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Khemlina G, Ikeda S, Kurzrock R. The biology of Hepatocellular carcinoma: implications for genomic and immune therapies. Mol Cancer. 2017;16(1):149. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Valero C, Lee M, Hoen D, et al. Pretreatment neutrophil‐to‐lymphocyte ratio and mutational burden as biomarkers of tumor response to immune checkpoint inhibitors. Nat Commun. 2021;12(1):729. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Chen S, Huang Z, Jia W, et al. Association of the pretreatment lung immune prognostic index with survival outcomes in advanced hepatocellular carcinoma patients treated with PD‐1 inhibitors. J Hepatocell Carcinoma. 2020;7:289‐299. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Le Cao KA, Meugnier E, McLachlan GJ. Integrative mixture of experts to combine clinical factors and gene markers. Bioinformatics. 2010;26(9):1192‐1198. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144(5):646‐674. [DOI] [PubMed] [Google Scholar]
- 21. Maccio A, Madeddu C. Blocking inflammation to improve immunotherapy of advanced cancer. Immunology. 2020;159(4):357‐364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Hill M, Segovia M, Russo S, Girotti MR, Rabinovich GA. The paradoxical roles of inflammation during PD‐1 blockade in cancer. Trends Immunol. 2020;41(11):982‐993. [DOI] [PubMed] [Google Scholar]
- 23. Ostroumov D, Fekete‐Drimusz N, Saborowski M, Kuhnel F, Woller N. CD4 and CD8 T lymphocyte interplay in controlling tumor growth. Cell Mol Life Sci. 2018;75(4):689‐713. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Lu J, Xu BB, Zheng ZF, et al. CRP/prealbumin, a novel inflammatory index for predicting recurrence after radical resection in gastric cancer patients: post hoc analysis of a randomized phase III trial. Gastric Cancer. 2019;22(3):536‐545. [DOI] [PubMed] [Google Scholar]
- 25. Luvian‐Morales J, Gonzalez‐Trejo S, Carrillo JF, et al. Association of the prognostic nutritional index and overall survival in patients with colorectal cancer: a STROBE compliant retrospective cohort study. Cancer Med. 2019;8(7):3379‐3388. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Rhodes B, Furnrohr BG, Vyse TJ. C‐reactive protein in rheumatology: biology and genetics. Nat Rev Rheumatol. 2011;7(5):282‐289. [DOI] [PubMed] [Google Scholar]
- 27. Pang RW, Poon RT. Diagnosis: novel prognostic biomarkers in hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol. 2012;9(12):691‐692. [DOI] [PubMed] [Google Scholar]
- 28. Jang JW, Oh BS, Kwon JH, et al. Serum interleukin‐6 and C‐reactive protein as a prognostic indicator in hepatocellular carcinoma. Cytokine. 2012;60(3):686‐693. [DOI] [PubMed] [Google Scholar]
- 29. Gehring AJ, Ho ZZ, Tan AT, et al. Profile of tumor antigen‐specific CD8 T cells in patients with hepatitis B virus‐related hepatocellular carcinoma. Gastroenterology. 2009;137(2):682‐690. [DOI] [PubMed] [Google Scholar]
- 30. Chraa D, Naim A, Olive D, Badou A. T lymphocyte subsets in cancer immunity: friends or foes. J Leukoc Biol. 2019;105(2):243‐255. [DOI] [PubMed] [Google Scholar]
- 31. Jiang Y, Tu X, Zhang X, et al. Nutrition and metabolism status alteration in advanced hepatocellular carcinoma patients treated with anti‐PD‐1 immunotherapy. Support Care Cancer. 2020;28(11):5569‐5579. [DOI] [PubMed] [Google Scholar]
- 32. Nakanishi Y, Masuda T, Yamaguchi K, et al. Albumin‐globulin ratio is a predictive biomarker of antitumor effect of anti‐PD‐1 antibody in patients with non‐small cell lung cancer. Int J Clin Oncol. 2020;25(1):74‐81. [DOI] [PubMed] [Google Scholar]
- 33. Marrelli D, Caruso S, Pedrazzani C, et al. CA19‐9 serum levels in obstructive jaundice: clinical value in benign and malignant conditions. Am J Surg. 2009;198(3):333‐339. [DOI] [PubMed] [Google Scholar]
- 34. Chen YL, Chen CH, Hu RH, Ho MC, Jeng YM. Elevated preoperative serum CA19‐9 levels in patients with hepatocellular carcinoma is associated with poor prognosis after resection. ScientificWorldJournal. 2013;2013:380797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Wan P, Zhang J, Long X, et al. Serum levels of preoperative alpha‐fetoprotein and CA19‐9 predict survival of hepatic carcinoma patients after liver transplantation. Eur J Gastroenterol Hepatol. 2014;26(5):553‐561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Hsu CC, Goyal A, Iuga A, et al. Elevated CA19‐9 Is associated with increased mortality in a prospective cohort of hepatocellular carcinoma patients. Clin Transl Gastroenterol. 2015;6:e74. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Material
Data Availability Statement
Data sharing is not applicable to this article as no datasets were generated or analyzed during the study.