The advanced lung cancer inflammation index is an independent prognostic factor after surgical resection in patients with non-small-cell lung cancer

Abstract

OBJECTIVES

The usefulness of a recently developed advanced lung cancer inflammation index (ALI) has been reported in advanced non-small-cell lung cancer (NSCLC) and small-cell lung cancer. However, no previous studies have examined the prognostic significance of ALI in patients with operable NSCLC. Therefore, the aim of this study was to explore the relationship between ALI and the prognosis of resected NSCLC.

METHODS

Three hundred and forty-three patients with NSCLC who underwent surgery at our institution between 2008 and 2012 were included. The ALI score was calculated as body mass index × serum albumin/neutrophil to lymphocyte ratio. A Web-based software programme [Cutoff Finder (http://molpath.charite.de/cutoff/)] was used to determine the optimal cut-off value for ALI. The Kaplan–Meier methods and a multivariable Cox proportional hazards model were used to evaluate the potential prognostic factors.

RESULTS

The optimal cut-off value of ALI was defined as 37.66. The low-ALI group (ALI < 37.66) displayed more adverse clinical characteristics. Furthermore, compared with patients in the high-ALI group (ALI > 37.66), those in the low-ALI group had significantly poorer survival rates. On multivariable analysis, gender, histological diagnosis, pN status, serum carcinoembryonic antigen level, serum C-reactive protein level and ALI were associated independently with cancer-specific survival.

CONCLUSIONS

This study is the first to investigate whether ALI is useful for predicting postoperative survival in patients with NSCLC. Preoperative ALI might serve as a potentially clinically valuable marker of the prognosis for patients with operable NSCLC.

Advanced lung cancer inflammation index, Non-small-cell lung cancer, Surgery, Prognosis

INTRODUCTION

A systemic inflammatory response has been well known to play a vital role in the progression of cancer, especially in advanced cases [1]. In addition, there is increasing evidence that elevated inflammatory parameters can be used for prognostication in operable cancers [2–7]. Previous investigations in patients with non-small-cell lung cancer (NSCLC) have shown that systemic inflammation, indicated by an elevated serum C-reactive protein (CRP) level and/or neutrophil to lymphocyte ratio (NLR), calculated as a simple ratio between neutrophil and lymphocyte counts, strongly influenced the prognosis after surgical resection [2–6]. Moreover, the Glasgow prognostic score, which is based on serum CRP and hypoalbuminaemia levels, has also been demonstrated as a prognostic factor in resected NSCLC [7–9].

Jafri et al. [10] evaluated a novel inflammation-based prognostic system, the advanced lung cancer inflammation index [ALI; based on body mass index (BMI), serum albumin (ALB) and NLR] in patients with metastatic NSCLC. Subsequently, the usefulness of ALI was reported in oesophageal cancer [11], small-cell lung cancer [12, 13] and diffuse large B-cell lymphoma [14].

To the best of our knowledge, the prognostic significance of ALI in operable NSCLC has not yet been investigated. Therefore, in this study, we reviewed our patients with NSCLC to identify the prognostic significance of ALI in patients with NSCLC who had surgery.

PATIENTS AND METHODS

This retrospective study was approved by our institutional review board, and the need to obtain patient consent was waived. A total of 391 consecutive patients with NSCLC who underwent curative surgery from 2008 to 2012 in our hospital were enrolled in the retrospective study. The blood samples for this analysis were collected before the operation. All the patients included in this study met the following criteria: (i) available preoperative NLR, serum ALB and CRP levels; (ii) adequate clinicopathological and follow-up data; (iii) no clinical evidence of infection or other inflammatory conditions and (iv) no second malignancies. The following patients were excluded: 33 patients without available preoperative NLR, 1 patient with preoperative lung abscess, 10 patients who were lost to follow-up and 4 patients who died of other diseases. There were no patients with pneumonia and other inflammatory diseases, except for 1 patient with an abscess. Finally, the collected records of 343 consecutive patients with NSCLC were reviewed retrospectively. No patients received preoperative chemoradiotherapy, and there were no surgical deaths.

The preoperative BMI was calculated as weight in kilograms divided by height in square metres. The ALI was calculated as follows: ALI = BMI × ALB (g/dl)/NLR [8]. A Web-based R software (Cutoff Finder) was used to determine the optimal cut-off value for ALI [12]. Cutoff Finder is a freely available Web application that can be accessed using an arbitrary Web browser (http://molpath.charite.de/cutoff) [15]. The preoperative serum carcinoembryonic antigen and cytokeratin 19 fragment (CYFRA 21-1) levels were also measured in all patients by enzyme immunoassay in a single laboratory at our hospital. The median postoperative follow-up was 62.5 months.

The χ² test was used to determine the association among the ALI groups. Cumulative cancer-specific survival curves after surgery were calculated using the Kaplan–Meier method, and differences were evaluated using the log-rank test. The Cox proportional hazard model was performed in univariable and multivariable analyses, and hazard ratios estimated from the Cox analysis are reported as relative risks with corresponding 95% confidence intervals. Statistical analyses were performed using JMP (SAS Institute Inc., Cary, NC, USA). A P-value of <0.05 indicated statistical significance.

RESULTS

There were 50 patients with BMI <18.5 kg/m², 200 with BMI = 18.5–24 kg/m² and 93 with BMI >24 kg/m². Using the biostatistical tool Cutoff Finder [15], we learned that a wide range of cut-off points for ALI were significant (259 of 314 tests, 82.5%; Fig. 1). The optimal cut-off point of ALI was determined to be 37.66. On the basis of this cut-off value, patients were subdivided into low ALI (ALI < 37.66) and high ALI (ALI > 37.66) groups. There were 121 (35.28%) patients in the low-ALI group; the remaining 222 (64.72%) patients were in the high-ALI group.

Figure 1

Hazard ratio for overall cancer-specific survival based on the cut-off point for the ALI in operated patients. The vertical line designates the optimal cut-off point with the most significant (log-rank test) split. The plot was generated using the biostatistical tool, Cutoff Finder. ALI: advanced lung cancer inflammation index; CI: confidence interval; OR: odds ratio.

The relationships between the ALI and the clinicopathological characteristics of the patients are listed in Table 1. Gender, smoking status, histological subtype, pStage, pN status, serum CYFRA21-1 level and serum CRP level were related to ALI. In other words, the low-ALI group displayed more adverse clinical characteristics. The mean CRP value in the low- and the high-ALI groups was 0.797 ± 1.681 and 0.145 ± 0.281, respectively (P < 0.001).

Table 1

Clinical characteristics of patients

	ALI		P-value
	Low	High
Age (years)
<65	37	69	0.923
≥65	84	153
Gender
Male	74	101	0.005
Female	47	121
Smoking status
Never	43	115	0.004
Current/former	78	107
Histological diagnosis
Adenocarcinoma	83	187	0.001
Other	38	35
pStage
I	84	182	0.009
II–III	37	40
pT status
pT1	77	160	0.108
pT2–3	44	62
pN status
pN0	96	197	0.021
pN1–2	25	25
CEA
Normal	88	168	0.550
High	33	54
CYFRA21-1
Normal	90	195	0.002
High	31	27
CRP
Low	70	174	<0.001
High	51	48

ALI: advanced lung cancer inflammation index; CEA: carcinoembryonic antigen; CRP: C-reactive protein; CYFRA21-1: cytokeratin 19 fragment.

Comparison of postoperative cancer-specific survival curves according to the ALI group showed a significant difference in the rates of patient survival (Fig. 2). The low ALI group had a significantly worse survival rate than did the high ALI group (P < 0.001).

Figure 2

Survival rate of patients based on the ALI. ALI: advanced lung cancer inflammation index.

In univariable analysis of overall survival, male gender (P < 0.001), current/former smoking status (P < 0.001), non-adenocarcinoma histological diagnosis (P < 0.001), pT2–3 status (P < 0.001), pN1–2 status (P < 0.001), high serum carcinoembryonic antigen level (P < 0.001), high serum CYFRA21-1 level (P < 0.0001), high serum CRP level (P < 0.001) and low ALI (P < 0.001) were significant factors for poor survival (Table 2).

Table 2

Univariable analysis

	Favourable	Unfavourable	Hazard ratio	P-value	95% Confidence interval
					Lower	Upper
Age (years)	<65	≥65	0.763	0.252	0.465	1.205
Gender	Female	Male	0.331	<0.001	0.205	0.518
Smoking	Never	Current/former	0.351	<0.001	0.216	0.552
Histological diagnosis	Adenocarcinoma	Other	0.249	<0.001	0.165	0.380
pT status	pT1	pT2–3	0.447	<0.001	0.296	0.678
pN status	pN0	pN1–2	0.344	<0.001	0.220	0.556
CEA	Normal	High	0.375	<0.001	0.247	0.575
CYFRA21-1	Normal	High	0.349	<0.001	0.226	0.550
CRP	Low	High	0.391	<0.001	0.259	0.591
ALI	High	Low	0.315	<0.001	0.206	0.477

ALI: advanced lung cancer inflammation index; CEA: carcinoembryonic antigen; CRP: C-reactive protein; CYFRA21-1: cytokeratin 19 fragment.

All significant prognostic factors tested via multivariable analysis were evaluated using the Cox proportional hazards model. In multivariable analyses, male gender (P = 0.003), non-adenocarcinoma histological diagnosis (P = 0.001), pN1–2 status (P = 0.002), high serum carcinoembryonic antigen level (P < 0.001), high serum CRP level (P = 0.046) and low ALI (P = 0.008) were independent predictors of poor prognosis (Table 3). ALI was an independent prognostic factor in operable NSCLC.

Table 3

Multivariable analysis

	Favourable	Unfavourable	Hazard ratio	P-value	95% Confidence interval
					Lower	Upper
Gender	Female	Male	0.362	0.003	0.179	0.711
Smoking	Never	Current/former	0.670	0.287	0.712	3.056
Histological diagnosis	Adenocarcinoma	Other	0.438	0.001	0.271	0.712
pT status	pT1	pT2–3	1.014	0.956	0.626	1.662
pN status	pN0	pN1–2	0.423	0.002	0.253	0.723
CEA	Normal	High	0.419	<0.001	0.267	0.664
CYFRA21-1	Normal	High	0.809	0.420	0.490	1.364
CRP	Low	High	0.631	0.046	0.403	0.993
ALI	High	Low	0.436	<0.001	0.278	0.679

ALI: advanced lung cancer inflammation index; CEA: carcinoembryonic antigen; CRP: C-reactive protein; CYFRA21-1: cytokeratin 19 fragment.

DISCUSSION

A growing body of evidence showed a connection between inflammation and cancer [1], and mechanistic studies have presented solid evidence to support the biological and prognostic importance of a pro-inflammatory tumour microenvironment in cancer progression [1].

NLR has been related to poor survival for operable NSCLC [3, 4]. Two meta-analyses also demonstrated that an elevated NLR was a predictor of poor overall survival in patients with lung cancer [5, 6]. The exact underlying mechanisms for the association of elevated NLR with the prediction of poor survival in patients with NSCLC have not been determined. An elevated NLR implies an increased neutrophil count and/or a decreased lymphocyte count as well as relative lymphopenia. A decreased lymphocyte count has been demonstrated as a biomarker of poor survival for patients with terminal cancer due to the key role of lymphocytes in killing cancer cells and regulating the proliferation, apoptosis, angiogenesis and metastasis of cancer by secreting cytokines [16, 17]. Therefore, NLR might serve as a host marker of tumour defence, which inhibits tumour cell proliferation and migration.

Nutritional status, generally evaluated using BMI, has also been closely linked to survival after surgery for NSCLC [17, 18]. Tewari et al. [18] and Nakagawa et al. [19] reported that, in patients with NSCLC, low BMI, a surrogate for impaired nutrition, was a negative predictor of long-term survival in resected NSCLC. ALB is also a good index for indicating nutritional status, and malnutrition has been reported to be associated with survival in patients having surgery for NSCLC [7–9, 20]. The underlying mechanism of this effect is unknown in detail, but impaired immune function due to nutritional depletion might lead to the inability of the body’s immune system to remove the cancer. Malnutrition and inflammation have been reported to suppress albumin synthesis [21]. Therefore, hypoalbuminaemia might serve as a clinically valuable marker of inflammation in addition to malnutrition in patients with operable NSCLC.

It is easy to understand that ALI, calculated from BMI, ALB and NLR, is a useful prognostic marker after surgical resection in patients with NSCLC. To the best of our knowledge, this study is the first to investigate whether ALI is useful for predicting the postoperative outcome in patients with NSCLC. Our results showed that preoperative ALI might serve as a potential clinical prognostic marker in patients with operable NSCLC.

Furthermore, our results also showed that low ALI was related to male gender, current/former smoking status, non-adenocarcinoma subtype, advanced pStage, pN1–2 status, high serum CYFRA21-1 level and high serum CRP level. Thus, decreased ALI might correlate with a more aggressive disease phenotype. However, the results of multivariable analysis showed that the prognostic significance of ALI is independent of these adverse clinical characteristics. This observation might mean that ALI is a Tumor-Node-Metastasis (TNM)-independent prognostic factor, despite a relationship between low ALI and advanced stage of disease.

CRP is an important acute-phase response protein produced mainly by hepatocytes, whose levels rise in response to inflammation [22]. Previous investigations [2–4] also found that elevated CRP is associated with poor survival in patients with resected NSCLC. The causal relationship between CRP and the prognosis may be explained by an aggressive cancer that results in a detrimental inflammatory reaction, thereby leading to elevated serum CRP levels [2]. Both NLR and CRP are markers of inflammation. However, in the present results, both NLR-based ALI and CRP were independent prognostic factors in multivariable analysis. Therefore, the underlying prognostic mechanisms of NLR and CRP might be different, at least in part.

ALI has the advantage of being simple to measure, routinely available and well standardized. Therefore, we believe that preoperative ALI is a proven, objective, reproducible, inexpensive predictor of survival in patients with NSCLC following surgical resection and should be considered for routine clinical use. Although operable patients with NSCLC with low ALI had poorer prognoses, we do not think that these patients should not receive surgery. These patients might be a suitable population for adjuvant chemotherapy.

Although this study shows the strong independent prognostic value of the ALI in operable patients with NSCLC, the retrospective nature and the relatively small sample size from a single centre should be acknowledged as potential limitations. Further large-scale population-based prospective studies are needed to fully consolidate the results.

CONCLUSIONS

In conclusion, low ALI could identify patients with poor prognosis and may be a useful marker in operable patients with NSCLC.

Conflict of interest: none declared.