Journal of Cancer Therapy, 2012, 3, 412-423
http://dx.doi.org/10.4236/jct.2012.324054 Published Online September 2012 (http://www.SciRP.org/journal/jct)
Clinical Biomarkers and Prognosis in Taiwanese Patients
with Non-Small Cell Lung Cancer (NSCLC)
Yixia Li1, Yea-Jyh Chen2, Li-Jung Chang3, Michael Hendryx1,4, Juhua Luo1,4,5*
1Department of Community Medicine, School of Medicine, West Virginia University, Morgantown, USA; 2School of Nursing, West
Virginia University, Morgantown, USA; 3Department of Nursing, Tzu Chi College of Technology, Hualien, Taiwan; 4West Virginia
Rural Health Research Center, Morgantown, USA; 5Mary Babb Randolph Cancer Center, West Virginia University, Morgantown,
USA.
Received May 9th, 2012; revised June 11th, 2012; accepted June 30th, 2012
ABSTRACT
Introduction: Lung cancer is the leading cause of cancer death worldwide with poor survival rates. However, the prog-
nostic factors for survival of patients with lung cancer are not well-established. In this study, we examined the impact of
routine laboratory biomarkers and traditional factors on survival of patients with non-small cell lung cancer (NSCLC).
Method: Secondary data analysis was conducted from a retrospective study of 404 patients with newly diagnosed lung
cancer in 2005-2007 in Taiwan. There were eight routine laboratory biomarkers and eight traditional factors investi-
gated in the analyses. Cox proportional hazards model was used to assess the hazard ratios for the association between
risk factors and patient overall survival. The Kaplan-Meier method was used to compare survival curves for each prog-
nostic indicator. Results: High WBC counts (HR = 1.798, 95%CI: 1.225 - 2.639), low Hgb level (HR = 1.437, 95%CI:
1.085 - 1.903), and low serum albumin level (HR = 2.049, 95%CI: 1.376 - 3.052) were significant laboratory prognostic
biomarkers for poor NSCLC survival. Additionally we confirmed the traditional prognostic factors for poor overall sur-
vival among NSCLC patients, including older age, comorbidity conditions, advanced cancer stage, and non-surgical
treatment. Conclusions: This study identified three available laboratory biomarkers, high WBC counts, low Hgb level,
and low serum albumin level, to be significant prognostic factors for poorer overall survival in NSCLC patients. Further
prognostic evaluation studies are warranted to compare different ethnic groups on the prognostic values of these clinical
parameters in NSCLC survival outcomes. These identified prognostic biomarkers should be included in early risk
screening of hospitalized lung cancer patient population.
Keywords: Clinical Biomarkers; Non-Small Cell Lung Cancer (NSCLC); Prognostic Factors
1. Introduction
Lung cancer is the leading cause of cancer death world-
wide [1]. An estimated 1.6 million new cases of lung
cancer occurred worldwide in 2008, accounting for about
13% of total cancer diagnoses [2]; and an estimated
951,000 men and 427,400 women died from the disease
worldwide in 2008. Histologically, lung cancer can be
classified into two types: small cell lung cancer (SCLC)
and non-small cell lung cancer (NSCLC). NSCLC ac-
counts for 85% of lung cancer cases, and approximately
65% of patients with NSCLC present with advanced-
stage (III or IV) disease [3]. In addition, lung cancer is
predominantly a disease of elderly people: the median
age of newly diagnosed lung cancer patients is approxi-
mately 68 years and as many as 40% of patients are older
than 70 years [4]. The five-year survival rates among
patients with NSCLC after complete resection are 46% -
81% with stage IA to IIB, compared to 21% - 24% with
stage IIIA to IV [5]. Furthermore, overall survival among
patients receiving chemotherapy also remains poor with a
median 4 to 15 monthssurvival time [6]. Therefore, study-
ing prognostic biomarkersas survival predictors may con-
tribute toimproveddisease prognosis and treatment guide-
lines for lung cancer.
Prognostic predictors for lung cancer outcomes are
inconsistent in previous literature. The most widely ac-
cepted prognostic factors for patients with NSCLC in-
clude tumor stage, performance status, and weight loss
[7-12]. Besides these factors, biochemical or hematologic
markers, such as white blood cell (WBC) counts, neu-
trophil counts and hemoglobin (Hgb) level, are also as-
sociated with NSCLC survival. A previous study per-
formed by the North Central Cancer Treatment Group
(NCCTG), which included 1053 lung cancer patients,
*Corresponding author.

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 413
revealed that patients who had high WBC counts and low
Hgb levels had significantly worse survival than their
counterparts [13]. However, the prognostic importance of
complete blood count findings were inconsistent with
other studies [14-18] where no significant association
was found for WBC counts or Hgb level. In addition,
neutrophil counts in a study by Ferrigno et al. were found
to independently predict the survival of NSCLC patients
[19]. Similar to Ferrigno’s result, the European Lung
Cancer Working Group found that a high-neutrophil
count was an independent prognostic factor for poor sur-
vival in patients with unresectable advanced NSCLC [20].
Paddisonalso reported that high level neutrophil counts
and low Hgblevel predicted poor survival of NSCLC
patients [21]. Furthermore, another studyfound that plate-
let countswere associated with poor overall survival in
aunivariate model but werenot significant in multivari-
ate analysis [22].
There may be additional biochemical markers related
to lung cancer survival, such as serum albumin level and
serum sodium level. Serum albumin level, a nutritional
indicator, has been examined as a prognostic lung cancer
marker among both SCLC and NSCLC patients in se-
veral previous studies [12,23,24]. Maeda et al. observed
that serum albumin level was one of the most significant
prognostic factors for advanced NSCLC in both univari-
ate and multivariate analysis [12]. In Win’s study, serum
albumin level was identified to be a significant prognos-
tic factor in univariatemodelbut not inmultivariable ana-
lysis for lung cancer patients [23]. A similar association
between serum albumin level and lung cancer survival
was observed by Tas et al. [24]. In addition, abnormal
serum sodium level has been studied asa poor survival
predictor for lung cancer. Previous research has reported
thathyponatremia predicts poor survival of SCLC pa-
tients [15,25-27], but there appears to be no previous
research onhyponatremiain relation to NSCLC survival
outcome.
Because of incomplete and in some cases inconsistent
evidence on prognostic biomarkers for NSCLC survival,
the aim of this study was to examine the relationships
between routine laboratory parameters including WBC
counts, neutrophil percentage, lymphocyte percentage,
Hgblevel, platelet counts, serum albumin level, and se-
rum sodium level on overall survival in NSCLC patients.
2. Materials and Methods
2.1. Study Design and Measures
This study is a secondary data analysis derived froma
retrospective study of patients with newly diagnosed lung
cancer in the years 2005-2007 in Taiwan. Detailed de-
scription of this study sample design and methods are
available in Luo et al. [28]. In brief, the retrospective
study screened potential patients through hospital medi-
cal records. Eligible patients were selected based on ini-
tial hospitalization with lung cancer as the primary diag-
nosis, who were discharged to home and were older than
40 years old. There were 404 patients eligible in the
study for final analysis.
Patients’ characteristics, extracted from hospital re-
cords, include age, gender, the Eastern Cooperative On-
cology Group (ECOG) performance (PS) (0 - 5 with 0
denoting perfect health and 5 denoting death) [29], smok-
ing status (never, former, current), body mass index
(BMI; kg/m2), Charlson comorbidity Index (CCI; ex-
cluded lung cancer as the primary disease), tumor stage
(less advanced—stage I, II and IIIA and advanced—stage
IIIB and IV), cancer treatment (surgery with/without one
or more types of supportive care [e.g. chemotherapy,
radiation, tyrosin kinase inhibitor targeted therapy], che-
motherapy with/without other supportive care, or no can-
cer treatment). The CCI, a weighted index score with a
possible range of 0 - 35, wasused to evaluate patients’
comorbid conditions according to the influence of co-
morbidity on overall mortality risk [30,31]. The patients’
BMI was calculated and classified into four categories:
underweight (<18.5 kg/m2), normal (18.5 - <25 kg/m2),
overweight (25 - <30 kg/m2) and obese (>=30 kg/m2).
Laboratory data including WBC counts, Hgb level, neu-
trophil percentage, lymphocyte percentage, platelet counts,
monocyte percentage, serum albumin level, and serum
sodium level were measured at the index visit. These
variables were considered as the main prognostic factors
of interest for the study. We used the admission or
first-time measure if more than one set of laboratory bio-
markers were examined during the index visit. In order
not to lose study power, we kept patients with missing
data in the final analysis and categorized those patients as
an additional group. The outcome of interest in this study
was survival duration, which was defined as the interval
in months from the discharge date of the index visit to
death from any cause. Patients’ survival data were follow-
ed through cancer registry records until the end of 2010.
2.2. Statistical Analysis
We categorized all potential prognostic variables includ-
ing WBC counts, Hgb level, neutrophil percentage, lym-
phocyte percentage, platelet counts, monocyte percentage,
albumin level, and sodium level into normal and abnor-
mal values according to standard laboratory norms (cut-
points are listed in Table 1) [32]. For sodium level, we
divided it into two groups since therewas only one pa-
tient with sodium level higher than 145 (mmol/l).
For univariate analyses, we estimated survival curves
using the Kaplan-Meier method and compared these
curves by the log-rank test. Survival duration was also
estimated by fitting the data with a Cox regression model
[33]. All variables reaching statistical significance (p <

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 414
Table 1. Characteristics of lung cancer patients during the study follow-up (N = 404).
Variable No. (%) Death within follow-up (N) Death within follow-up (%)
Age (years)
<70 201(49.75) 167 83.08
≥70 203 (50.25) 185 91.13
Gender
Female 148 (36.63) 124 83.78
Male 256 (63.37) 228 89.06
Smoking status
Never 179 (44.31) 149 83.24
Former 156 (38.61) 141 90.38
Current 69 (17.08) 62 89.86
ECOG Performance status
<2 264 (65.35) 217 82.20
≥2 140 (34.65) 135 96.43
BMI (kg/m2)
<18.5 39 (9.65) 36 92.31
18.5 - <25 225 (55.69) 193 85.78
25-<30 86 (21.29) 70 81.40
≥30 13 (3.22) 12 92.31
Missing 41 (10.15) 41 100.00
Co-morbidity score
<4 136 (33.66) 105 77.21
4 - 8 200 (49.50) 184 92.00
≥8 68 (16.83) 63 92.65
Cancer stage
Less advanced 41 (10.15) 25 60.98
Advanced 363 (89.85) 327 90.08
Cancer treatment
Surgery 24 (5.94) 8 33.33
Chemotherapy and /or other supportive care 300 (74.26) 268 89.33
None 80 (19.80) 76 95.00
WBC counts (103/µL)
≤5.5 63 (15.59) 52 82.54
5.5 - 15.5 300 (74.26) 260 86.67
>15.5 41 (10.15) 40 97.56
Neutrophil (%)
27 - 55 35 (8.66) 25 71.43
>55 339 (83.91) 299 88.20
Missing 30 (7.43) 28 93.33
Lymphocyte (%)
<16 175 (43.32) 162 92.57
16 - 46 220 (54.46) 187 85.00
>46 9 (2.23) 3 33.33

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 415
Continued
Hemoglobin (g/dL)
<11.5 122 (30.20) 118 96.72
11.5 - 13.5 152 (37.62) 129 84.87
>13.5 130 (32.18) 105 80.77
Platelet counts (Thou/μl)
<150 157(38.86) 133 84.71
150 - 400 164 (40.59) 144 87.80
>400 51 (12.62) 48 94.12
Missing 32 (7.92) 27 84.38
Monocyte (%)
<4 99 (24.50) 85 85.86
4 - 11 280 (69.31) 243 86.79
>11 25 (6.19) 24 96.00
Albumin (g/dl)
<3.1 40 (9.90) 40 100.00
3.1 - 4.3 175 (43.32) 151 86.29
>4.3 28 (6.93) 20 71.43
missing 161 (39.85) 141 87.58
Sodium (mmol/l)
<136 189 (46.78) 177 93.65
136 - 147 215 (53.22) 175 81.40
0.05) at the univariate level were included in the multi-
variate analysis. We performed Pearson correlations
among the selected variables for collinearity before the
multivariate analysis. If any two variables were highly
correlated (r > 0.6), only one was selected in the multi-
variate model. We used the Cox proportional hazards
model to perform the multivariate analysis to adjust for
all included variables. The hazard ratios were calculated
to assess the death risk for various prognostic factors. All
statistical analyses were done using SAS version 9.2 with
a significant p-value criterion of 0.05 or less.
3. Results
3.1. Baseline Patient Characteristics
Of the total 404 patients with newly diagnosed NSCLC,
the median follow-up duration time was 10.7 months
(range, 0 - 70 months). Mean age of the study sample
population was 67.6 years (SD = 11.0). Over 60% of the
patients (N = 256) were male; approximately 39% of pa-
tients were former smokers and 17% were current smok-
ers. The majority of patients (90%) were diagnosed at an
advanced stage (stage IIIB or stage IV). Overall, 352
patients (87%) died by the end of 2010. Table 1 summa-
rizes the study patient characteristics.
3.2. Univariate Analysis
Univariate analysis revealed the following patient char-
acteristics to be significant prognostic factors for poor
survival: older age (≥70 years), male, current smoker,
poor performance status (PS ≥ 2), low BMI (≤18.5
kg/m2), high CCI score (≥4), advanced cancer stage (IIIB
and IV), cancer treatment (chemotherapy with/without
other supportive cancer treatment or no cancer treatment),
high WBC (>15.5 × 103/µL), high neutrophil percentage
(>55%), lower (<16%)/ higher (>46%) lymphocyte per-
centage, low Hgblevel (<11.5 g/dl), lower (<3.1 g/dl)/
higher (>4.3 g/dl) albumin level, and low sodium level
(<136 mmol/l). Other variables, including platelet counts
and monocyte percentage were not observed to be sig-
nificantly associated with overall NSCLC survival (p >
0.05; see Table 2).
In patients with advanced cancer stage, the median
survival time was 10.1 (95%CI: 8.6 - 12.0) months com-
pared with 26.2 survival months among those patients
with less advanced stage (p < 0.0001). In patients with
high WBC counts, the median survival was 1.93 (95%CI:
1.1 - 4.1) months, compared with the patients with nor-
mal and low WBC counts (11.9 vs 14.3 survival months,
respectively; p < 0.0001). Similarly, patients with low

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 416
Table 2. Univariate analyses of prognostic factors and survival in patients with NSCLC (N = 404).
Variable Hazard ratio 95%CI p-value
Age (years)
<70 1
≥70 1.703 1.380 - 2.102 <0.0001
Gender
Female 1
Male 1.273 1.023 - 1.585 0.0306
Smoking status
Never 1
Former 1.263 0.939 - 1.699 0.1228
Current 1.410 1.119 - 1.777 0.0035
ECOG performance status
<2 1
≥2 2.236 1.798 - 2.781 <0.0001
BMI (kg/m2)
<18.5 1.430 1.002 - 2.043 0.0490
18.5 - <25 1
25 - <30 0.969 0.541 - 1.738 0.9166
≥30 0.949 0.722 - 1.247 0.7068
missing 4.013 2.838 - 5.674 <0.0001
Co-morbidity score
<4 1
4 - 8 1.479 1.162 - 1.883 0.0015
≥8 1.873 1.366 - 2.568 <0.0001
Cancer stage
Less advanced 1
Advanced 2.375 1.575 - 3.582 <0.0001
Cancer treatment
Surgery 1
Chemotherapy and/ or other supportive care 4.877 2.408 - 9.877 <0.0001
None 11.481 5.522 - 23.868 <0.0001
WBC counts (103/µL)
≤ 5.5 0.789 0.586 - 1.063 0.1192
5.5 - 15.5 1
>15.5 2.642 1.889 - 3.697 <0.0001
Neutrophil (%)
27 - 55 1
>55 1.942 1.290 - 2.923 0.0015
Missing 1.897 1.105 - 3.257 0.0203
Lymphocyte (%)
<16 1.620 1.311 - 2.003 <0.0001
16 - 46 1
>46 0.240 0.077 - 0.752 0.0143

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 417
Contunued
Hemoglobin (g/dL)
<11.5 1.908 1.482 - 2.455 <0.0001
11.5 - 13.5 1
>13.5 0.786 0.607 - 1.018 0.0679
Platelet counts (Thou/μl)
<150 0.803 0.634 - 1.016 0.6628
150 - 400 1
>400 1.295 0.933 - 1.797 0.1227
Missing 0.913 0.605 - 1.377 0.6628
Monocyte (%)
<4 0.858 0.670 - 1.099 0.2248
4 - 11 1
>11 1.310 0.861 - 1.993 0.2080
Albumin (g/dl)
<3.1 3.728 2.598 - 5.350 <0.0001
3.1 - 4.3 1
>4.3 0.609 0.381 - 0.972 0.0374
Missing 1.070 0.850 - 1.346 0.5657
Sodium (mmol/l)
<136 1.783 1.443 - 2.201 <0.0001
136 - 147 1
Hgb level (<11.5 g/dL) had 5.4 (95%CI: 4.1 - 7.2) sur-
vival months, compared with 12.2 months for patients
with normal Hgb level and 7.5 months for patients with
high Hgb level. For patients with low albumin level, the
overall survival time was shorter (2.4 months; 95%CI:
1.2 - 5.9), compared with those having normal and high
albumin levels (12.4 months vs 22.0 months, respectively;
p < 0.0001). Survival curvesfor these four markers are
shown in Figures 1(a)-(d).
3.3. Multivariate Analysis
After performing the Pearson correlations, gender and
smoking status were highly correlated (r = 0.68). Gender
remained in the multivariate model for analysis because
1) it is a common characteristic for analysis; and 2) in an
analysis of males only, smoking status was not a signifi-
cant independent predictor for NSCLC (data not shown).
After simultaneously adjusting for all potential prognos-
tic factors (see Table 3), significantly lower survival
rates were identifiedfor patients with older age (≥70
years), lower performance status, higher co-morbidity
score (>8), advanced lung cancer stage, cancer treatment
rather than surgery, high WBC counts, low Hgblevels
and low albumin levels. Correspondingly, similar results
were found while performing a multivariable analysis
stratified by cancer stage (data not shown). However, our
data were insufficient to perform a similar analysis for
patients with less advanced cancer stage, as only 7.1%
patients with less advanced cancer diagnosis died within
the study period.
4. Discussion
Our study demonstrated that older age, poor performance
status, poor comorbid condition, advanced cancer stage,
treatment other than surgery, high WBC counts, low Hgb
levels and low albumin levels were significant prognostic
factors forshorter NSCLC survival in multivariate analy-
sis.
Tumor stage, performance status and weight loss at the
time of diagnosis have been found tohave a negative im-
pact on patient survival from NSCLC in previous re-
search [34,35]. The present study confirmed that ad-
vanced cancer stage and poor performance status were
prognostic factors for NSCLC survival. However, low
BMI (<18.5 kg/m2) was a significant prognostic factor in
univariate analysis but not in multivariate analysis in our
study. This may due to the fact that BMI data were miss-
ing in 10% of cases. We observed that patients with
missing BMI all died within the study follow-up period,
indicating that patients with unavailable BMI data during
hospitalization were more likely to be severely ill with
relatively higher mortality rate. In addition, individuals
with missing BMI data tended to be older, with advanced
cancer stage, and have poor performance status.

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 418
(a) (b)
(c) (d)
Figure 1. NSCLC survival curve by cancer staging and laboratory biomarkers (N = 404). (a) Survival curves by cancer stage:
1 for less advanced stage, 2 for advanced stage (p < 0.0001); (b) Survival curves by WBC counts: 1 for ≤ 5.5 × 103/µL, 2 for
5.5 × 103/µL - 15.5 × 103/µL, 3 for > 15.5 × 103/µL (p < 0.0001); (c) Survival curves by hemoglobin (Hgb) level: 1 for ≤11.5
g/dL, 2 for 11.5 g/dL - 13.5 g/dL, 3 for >13.5 g/dL (p < 0.0001); (d) Survival curves by serum albumin (alb) level: 1 for ≤ 3.1
g/dL, 2 for 3.1 g/dL - 4.3 g/dL, 3 for >4.3 g/dL, 4 for missing value (p < 0.0001).
Table 3. Multivariate analyses of key prognostic factors and survival in patients with NSCLC (N = 404).
Variable Hazard ratio 95%CI p-value
Age
<70 1
≥70 1.445 1.147 - 1.819 0.0017
Gender
Female 1
Male 1.183 0.936 - 1.494 0.1597
BMI (kg/m2)
<18.5 0.980 0.658 - 1.460 0.9207
18.5 - <25 1
25 - <30 0.998 0.750 - 1.329 0.9898
≥30 1.129 0.619 - 2.065 0.6924
missing 2.269 1.514 - 3.399 <0.0001

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 419
Contunued
ECOG Performance status
<2 1
≥2 1.506 1.167 - 1.943 0.0017
Co-morbidity score
<4 1
4 - 8 1.354 1.023 - 1.791 0.0340
≥8 1.557 1.103 - 2.199 0.0118
Cancer stage
Less advanced 1
Advanced 1.818 1.153 - 2.867 0.0101
Cancer treatment
Surgery 1
Chemotherapy and/ or other supportive care 2.513 1.202 - 5.257 0.0144
None 5.704 2.665 - 12.211 <0.0001
WBC counts (103/µL)
≤5.5 0.798 0.577 - 1.104 0.1724
5.5 - 15.5 1
>15.5 1.798 1.225 - 2.639 0.0027
Neutrophil (%)
27 - 55 1
>55 1.308 0.826 - 2.070 0.2521
Missing 1.650 0.891 - 3.055 0.1112
Lymphocyte (%)
<16 0.931 0.705 - 1.230 0.6157
16 - 46 1
>46 0.475 0.139 - 1.622 0.2350
Hemoglobin (g/dL)
<11.5 1.437 1.085 - 1.903 0.0115
11.5-13.5 1
>13.5 0.901 0.686 - 1.183 0.4522
Albumin (g/dl)
<3.1 2.149 1.376 - 3.052 0.0004
3.1 - 4.3 1
>4.3 0.943 0.580 - 1.534 0.8131
Missing 1.279 1.006 - 1.625 0.0443
Sodium (mmol/l)
<136 1.141 0.895 - 1.454 0.2875
136 - 147 1
Comorbidity condition was also observed as a predic-
tor of lung cancer survival. Previous findings on comor-
bidity are inconsistent. Janssen et al. reported no inde-
pendent prognostic effect of comorbidity condition for
patients with non-small cell lung cancer [36]. However,
in line with our study, Wang and colleagues revealed that
patients with CCI score ≥ 2 had higher perioperative
mortality and death from NSCLC compared with patients
with CCI score < 2 [37]. Firat et al. and Moro-Sibilot et
al. also studied the significance of comorbidity scores in
stage I NSCLC patients and found it to be a significant
prognostic impact [38,39].
Our study confirmed that elevated WBC counts were
significantly associated with poor prognosis among

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 420
NSCLC patients, which aligned with most previous ob-
servations [13]. In a nationally representative adult co-
hort study, WBC counts werefound to be positively and
independently associated with cancer mortality after ad-
justing for age, gender and race [40]. In that study, in-
flammation was associated with WBC counts as well.
WBC as a marker of inflammatory level reflects either a
greater burden of tumor cells within the bone marrow, a
possible concomitant subclinical infection, or the effect
of a yet undescribed chemokine or cytokine secreted by
the tumor into the circulation. Stromal tissues of tumors
contain large numbers of WBC counts, and the inflam-
matory cell number and their cytokine production corre-
late with tumor severity and prognosis [41-44].
We also identified the relationship of both low Hgb
and serum albumin levels with poor survival as reported
in other studies [12,13,23,24,45,46]. Takugawa et al.
stated low Hgb level correlated with overall survival
among patients with NSCLC [47]. Similar findings were
reported in Albain’s study [48]. For serum albumin level,
Phillips reported a marked increase in mortality rate with
decreasing serum albumin concentrations among cancer
and cardiovascular patients [49]. Another study found an
approximate 25% reduction in cancer mortality among
middle aged men with a one standard deviation increase
in serum albumin [50]. Similar to the elevated WBC
counts, low Hgb and serum albumin levels also play an
influential role in body inflammation. Serum albumin is a
negative acute phase protein; its concentration in the
blood is reduced in response to inflammation [51]. Hy-
poalbuminemia is the result of the combined effects of
inflammation and inadequate protein and caloric intake
in patients with chronic disease and cancer. Inflammation
and malnutrition both reduce albumin concentration by
decreasing its rate of synthesis, while inflammation alone
is associated with a greater fractional catabolic rate (FCR)
and, when extreme, increased transfer of albumin out of
the vascular compartment [50]. This study is the first to
assess clinical prognostic factors in a Taiwanese popula-
tion, although several other Asian research articles havei-
dentified the traditional factors for lung cancer survival
including old age, no surgery treatment, performance
status, and advanced lung cancer stage [28,52,53].
The findings of this study should be considered in the
context of its strengths and limitations. Study strengths
include the fact that 1) all lung cancer cases were newly
diagnosed which ruled out impact on patients’ outcomes
by possible cancer pretreatment that patients may have
received. In addition, 2) the outcome measure (death
yes/no) was tracked up to 3 - 5 years in order to predict
long term prognosis for NSCLC; and 3) we were able to
collect and adjust for most potential prognostic factors
such as smoking status, BMI, and performance status.
However, the present study was also limited in several
respects. Using data derived from retrospective hospital
records review, we found that routinely examined labo-
ratory item such as albumin level was not available for
every individual during the hospital stay. These missing
data limited the ability of our study to conduct a com-
plete data analysis. In addition, the study uses a sample
from a Taiwanese patient populationmost of whom reside
in a rural area, which may limit the generalizability of
study findings of NSCLC survival outcomes to other
populations.
5. Conclusion
This study contributes to research on overall NSCLC
survival by concurrently identifying significant labora-
tory biomarkers for survival, including WBC counts,
Hgb level, and serum albumin level. Additionally we
confirmed the widely accepted prognostic factors of lung
cancer survival such as old age, advanced cancer, severe
comorbidity, performance status, and lack of surgery
treatment. Because these identified biomarkers are rou-
tinelychecked during hospital admissions, the findings of
this study could help in the early identification of patients
atrisk of shorter NSCLC survival to provide better clini-
cal patients care. Moreover, to extend and confirm the
current study findings, future studies shouldapply more
comprehensive prognostic assessments, conduct longer
follow-ups, and study additional ethnic populations.
6. Acknowledgements
This study was supported by the Faculty Research Grant
(TCCT-981A11), Buddhist Tzu Chi College of Tech-
nology, to L. J. Chang. We would like to express our
gratitude to Tzu Chi General Hospital Dalin Branch and
Cancer Tumor Center for their efforts and assistance for
data collection.
REFERENCES
[1] A. Jemal, R. Siegel, E. Ward, Y. Hao, J. Xu and M. J.
Thun, “Cancer Statistics, 2009,” CA—A Cancer Journal
for Clinicians, Vol. 59, No. 4, 2009, pp. 225-249.
doi:10.3322/caac.20006
[2] A. Jemal, F. Bray, M. M. Center, J. Ferlay, E. Ward and
D. Forman, “Global Cancer Statistics,” CA—A Cancer
Journal for Clinicians, Vol. 61, No. 6, 2011, pp. 69-90.
doi:10.3322/caac.20107
[3] C. L. B. Ott, N. Ratna, R. Prayag, Z. Nugent, K. Badiani
and S. Navaratnam, “Survival and Treatment Patterns in
Elderly Patients with Advanced Non-Small-Cell Lung
Cancer in Manitoba,” Current Oncology, Vol. 18, No. 5,
2011, pp. e238-e242.
[4] E. Quoix, V. Westeel, G. Zalcman and B. Milleron,
“Chemotherapy in Elderly Patients with Advanced Non-
Small Cell Lung Cancer,” Lung Cancer, Vol. 74, No. 3,

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 421
2011, pp. 364-368. doi:10.1016/j.lungcan.2011.06.006
[5] W. D. Wei, Z. S. Wen, X. D. Su, P. Lin, T. H. Rong and
L. K. Chen, “Multivariate Survival Analysis of 899 Pa-
tients with Non-Small Cell Lung Cancer after Complete
Resection,” Cancer, Vol. 26, No. 11, 2007, pp. 1231-
1236.
[6] K. Hotta, Y. Fujiwara, K. Kiura, N. Takigawa, M. Tabata,
H. Ueoka and M. Tanimoto, “Relationship between Re-
sponse and Survival in More than 50,000 Patients with
Advanced Non-Small Cell Lung Cancer Treated with
Systemic Chemotherapy in 143 Phase III Trials,” Journal
of Thoracic Oncology, Vol. 2, 2007, pp. 402-407.
doi:10.1097/01.JTO.0000268673.95119.c7
[7] J. L. Port, M. S. Kent, R. J. Korst, D. Libby, M. Pasman-
tier and N. K. Altorki, “Tumor Size Predicts Survival
within Stage IA Non-Small Cell Lung Cancer,” Chest,
Vol. 124, No. 5, 2003, pp. 1828-1833.
doi:10.1378/chest.124.5.1828
[8] E. F. Patz Jr., S. Rossi, D. H. Harpole Jr., J. E. Herndon
and P. C. Goodman, “Correlation of Tumor Size and Sur-
vival in Patients with Stage IA Non-Small Cell Lung
Cancer,” Chest, Vol. 117, 2000, pp. 1568-1571.
doi:10.1378/chest.117.6.1568
[9] L. M. Forrest, D. C. McMillan, C. S. McArdle, W. J.
Angerson, K. Dagg and H. R. Scott, “A Prospective Lon-
gitudinal Study of Performance Status, an Inflamma-
tion-Based Score (GPS) and Survival in Patients with In-
operable Non-Small-Cell Lung Cancer,” British Journal
of Cancer, Vol. 92, 2005, pp. 1834-1836.
doi:10.1038/sj.bjc.6602591
[10] H. R. Scott, D. C. McMillan, L. M. Forrest, D. J. Brown,
C. S. McArdle and R. Milroy, “The Systemic Inflamma-
tory Response, Weight Loss, Performance Status and
Survival in Patients with Inoperable Non-Small Cell Lung
Cancer,” British Journal of Cancer, Vol. 87, No. 3, 2002,
pp. 264-267. doi:10.1038/sj.bjc.6600466
[11] M. K. Mohamed, S. Ramalingam, Y. Lin, W. Gooding
and C. P. Belani, “Skin Rash and Good Performance
Status Predict Improved Survival with Gefitinib in Pa-
tients with Advanced Non-Small Cell Lung Cancer,” An-
nals of Oncology, Vol. 16, No. 5, 2005, pp. 780-785.
doi:10.1093/annonc/mdi157
[12] T. Maeda, H. Ueoka, M. Tabata, K. Kiura, T. Shibayama,
K. Gemba, N. Takigawa, A. Hiraki, H. Katayama and M.
Harada, “Prognostic Factors in Advanced Non-Small Cell
Lung Cancer: Elevated Serum Levels of Neuron Specific
Enolase Indicate Poor Prognosis,” Japanese Journal of
Clinical Oncology, Vol. 30, No. 12, 2000, pp. 534-541.
doi:10.1093/jjco/hyd139
[13] S. J. Mandrekar, S. E. Schild, S. L. Hillman, K. L. Allen,
R. S. Marks, J. A. Mailliard, J. E. Krook, A. W. Mak-
symiuk, K. Chansky, K. Kelly, A. A. Adjei and J. R. Jett,
“A Prognostic Model for Advanced Stage Nonsmall Cell
Lung Cancer. Pooled Analysis of North Central Cancer
Treatment Group Trials,” Cancer, Vol. 107, No. 4, 2006,
pp. 781-792. doi:10.1002/cncr.22049
[14] C. W. Francis, A. Blinc, S. Lee and C. Cox, “Ultrasound
Accelerates Transport of Recombinant Tissue Plasmino-
gen Activator into Clots,” Ultrasound in Medicine & Bi-
ology, Vol. 21, No. 3, 1995, pp. 419-424.
doi:10.1016/0301-5629(94)00119-X
[15] L. R. Zacharski, M. Z. Wojtukiewicz, V. Costantini, D. L.
Ornstein and V. A. Memoli, “Pathways of Coagula-
tion/Fibrinolysis Activation in Malignancy,” Semin Thr-
omb Hemost, Vol. 18, No. 1, 1992, pp. 104-116.
doi:10.1055/s-2007-1002415
[16] K. R. Meehan, L. R. Zacharski, T. E. Moritz and F. R.
Rickles, “Pretreatment Fibrinogen Levels Are Associated
with Response to Chemotherapy in Patients with Small
Cell Carcinoma of the Lung: Department of Veterans Af-
fairs Cooperative Study 188,” American Journal of He-
matology, Vol. 49, 1995, pp. 143-148.
doi:10.1002/ajh.2830490208
[17] O. Taguchi, E. C. Gabazza, H. Yasui, T. Kobayashi, M.
Yoshida and H. Kobayashi, “Prognostic Significance of
Plasma D-Dimer Levels in Patients with Lung Cancer,”
Thorax, Vol. 52, No. 6, 1997, pp. 563-565.
doi:10.1136/thx.52.6.563
[18] M. Z. Wojtukiewicz, L. R. Zacharski, T. E. Moritz, K.
Hur, R. L. Edwards and F. R. Rickles, “Prognostic Sig-
nificance of Blood Coagulation Tests in Carcinoma of the
Lung and Colon,” Blood Coagul Fibrinolysis, Vol. 3, No.
4, 1992, pp. 429-437.
[19] D. Ferrigno, G. Buccheri and I. Ricca, “Prognostic Sig-
nificance of Blood Coagulation Tests in Lung Cancer,”
European Respiratory Journal, Vol. 17, No. 4, 2001, pp.
667-673. doi:10.1183/09031936.01.17406670
[20] M. Paesmans, J. P. Sculier, P. Libert, G. Bureau, G.
Dabouis, J. Thiriaux, J. Michel, O. Van Cutsem, R. Ser-
gysels, P. Momme, et al., “Prognostic Factors for Sur-
vival in Advanced Non-Small-Cell Lung Cancer: Uni-
variate and Multivariate Analyses Including Recursive
Partitioning and Amalgamation Algorithms in 1052 Pa-
tients. The European Lung Cancer Working Party,”
Journal of Clinical Oncology, Vol. 13, No. 5, 1995, pp.
1221-1230.
[21] J. S. Paddison, J. S. Temel, G. L. Fricchione and W. F.
Pirl, “Using the Differential from Complete Blood Counts
as a Biomarker of Fatigue in Advanced Non-Small-Cell
Lung Cancer: An Exploratory Analysis,” Palliat Support
Care, Vol. 7, No. 2, 2009, pp. 213-217.
doi:10.1017/S1478951509000273
[22] D. Ferrigno and G. Buccheri, “Hematologic Counts and
Clinical Correlates in 1201 Newly Diagnosed Lung Can-
cer Patients,” Monaldi Archives for Chest Disease, Vol.
59, No. 3, 2003, pp. 193-198.
[23] T. Win, L. Sharples, A. M. Groves, A. J. Ritchie, F. C.
Wells and C. M. Laroche, “Predicting Survival in Poten-
tially Curable Lung Cancer Patients,” Lung, Vol. 186, No.
2, 2008, pp. 97-102. doi:10.1007/s00408-007-9067-1
[24] F. Tas, A. Aydiner, E. Topuz, H. Camlica, P. Saip and Y.
Eralp, “Factors Influencing the Distribution of Metastases
and Survival in Extensive Disease Small Cell Lung Can-
cer,” Acta Oncologica, Vol. 38, No. 8, 1999, pp. 1011-
1015. doi:10.1080/028418699432275
[25] S. L. Vanhees, R. Paridaens and J. F. Vansteenkiste,
“Syndrome of Inappropriate Antidiuretic Hormone Asso-
ciated with Chemotherapy-Induced Tumour Lysis in

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 422
Small-Cell Lung Cancer: Case Report and Literature Re-
view,” Annals of Oncology, Vol. 11, 2000, pp. 1061-1065.
doi:10.1023/A:1008369932384
[26] O. Hansen, P. Sorensen and K. H. Hansen, “The Occur-
rence of Hyponatremia in SCLC and the Influence on
Prognosis: A Retrospective Study of 453 Patients Treated
in a single Institution in a 10-Year Period,” Lung Cancer,
Vol. 68, No. 1, 2010, pp. 111-114.
doi:10.1016/j.lungcan.2009.05.015
[27] N. S. Rawson and J. Peto, “An Overview of Prognostic
Factors in Small Cell Lung Cancer. A Report from the
Subcommittee for the Management of Lung Cancer of the
United Kingdom Coordinating Committee on Cancer Re-
search,” British Journal of Cancer, Vol. 61, No. 4, 1990,
pp. 597-604. doi:10.1038/bjc.1990.133
[28] J. Luo, Y. J. Chen and L. J. Chang, “Fasting Blood Glu-
cose Level and Prognosis in Non-Small Cell Lung Cancer
(NSCLC) Patients,” Lung Cancer, Vol. 76, No. 2, 2011,
pp. 242-247. doi:10.1016/j.lungcan.2011.10.019
[29] M. M. Oken, R. H. Creech, D. C. Tormey, J. Horton, T. E.
Davis, E. T. McFadden and P. P. Carbone, “Toxicity and
Response Criteria of the Eastern Cooperative Oncology
Group,” American Journal of Clinical Oncology, Vol. 5,
No. 6, 1982, pp. 649-655.
doi:10.1097/00000421-198212000-00014
[30] M. E. Charlson, P. Pompei, K. L. Ales and C. R.
MacKenzie, “A New Method of Classifying Prognostic
Comorbidity in Longitudinal Studies: Development and
Validation,” Journal of Chronic Diseases, Vol. 40, No. 5,
1987, pp. 373-383. doi:10.1016/0021-9681(87)90171-8
[31] O. Birim, A. P. Kappetein and A. J. Bogers, “Charlson
Comorbidity Index as a Predictor of Long-Term Outcome
after Surgery for Nonsmall Cell Lung Cancer,” European
Journal Cardio-Thoracic Surgery, Vol. 28, No. 5, 2005,
pp. 759-762. doi:10.1016/j.ejcts.2005.06.046
[32] J. V. Corbett, “Laboratory Tests and Diagnostic Proce-
dures with Nursing Diagnoses,” Pearson/Prentice Hall,
Upper Saddle River, 2004.
[33] D. R. Cox and D. Oakes, “Analysis of Survival Data,”
Chapman and Hall, London, 1984.
[34] S. Leij-Halfwerk, P. C. Dagnelie, J. W. van Den Berg, J.
D. Wattimena, C. H. Hordijk-Luijk and J. P. Wilson,
“Weight Loss and Elevated Gluconeogenesis from Alanine
in Lung Cancer Patients,” American Journal of Clinical
Nutrition, Vol. 71, 2000, pp. 583-589.
[35] H. Wertzel, H. Siebert, W. Lange, L. Swoboda, E. Graf
and J. Hasse, “Results after Surgery in Stage-I Broncho-
genic Carcinoma,” Journal of Thoracic and Cardiovas-
cular Surgery, Vol. 46, 1998, pp. 365-369.
doi:10.1055/s-2007-1010255
[36] R. R. Jennens, G. G. Giles and R. M. Fox, “Increasing
Under Representation of Elderly Patients with Advanced
Colorectal or Non-Small-Cell Lung Cancer in Chemo-
therapy Trials,” Internal Medicine Journal, Vol. 36, 2006,
pp. 216-220. doi:10.1111/j.1445-5994.2006.01033.x
[37] C. Y. Wang, Y. S. Lin, C. Tzao, H. C. Lee, M. H. Huang,
W. H. Hsu and H. S. Hsu, “Comparison of Charlson Co-
morbidity Index And Kaplan-Feinstein Index in Patients
with Stage I Lung Cancer after Surgical Resection,”
European Journal Cardio-Thoracic Surgery, Vol. 32, No.
6, 2007, pp. 877-881. doi:10.1016/j.ejcts.2007.09.008
[38] S. Firat, M. Bousamra, E. Gore and R. W. Byhardt, “Co-
morbidity and KPS Are Independent Prognostic Factors
in Stage I Non-Small-Cell Lung Cancer,” International
Journal of Radiation Oncology*Biology*Physics, Vol. 52,
2002, pp. 1047-1057.
[39] D. Moro-Sibilot, A. Aubert, S. Diab, S. Lantuejoul, P.
Fourneret, E. Brambilla, C. Brambilla and P. Y. Brichon,
“Comorbidities and Charlson Score in Resected Stage I
nonsmall Cell Lung Cancer,” European Respiratory
Journal, Vol. 26, No. 3, 2005, pp. 480-486.
doi:10.1183/09031936.05.00146004
[40] A. Shankar, J. J. Wang, E. Rochtchina, M. C. Yu, R.
Kefford and P. Mitchell, “Association between Circulat-
ing White Blood Cell Count and Cancer Mortality: A
Population-Based Cohort Study,” Archives of Internal
Medicine, Vol. 166, 2006, pp. 188-194.
doi:10.1001/archinte.166.2.188
[41] L. M. Coussens and Z. Werb, “Inflammation and Can-
cer,” Nature, Vol. 420, No. 6917, 2002, pp. 860-867.
doi:10.1038/nature01322
[42] F. Balkwill, “Cancer and the Chemokine Network,” Na-
ture Reviews Cancer, Vol. 4, 2004, pp. 540-550.
doi:10.1038/nrc1388
[43] F. Balkwill and L. M. Coussens, “Cancer: An Inflamma-
tory Link,” Nature, Vol. 431, No. 7007, 2004, pp. 405-
406. doi:10.1038/431405a
[44] E. Pikarsky, R. M. Porat, I. Stein, R. Abramovitch, S.
Amit, S. Kasem, E. Gutkovich-Pyest, S. Urieli-Shoval, E.
Galun and Y. Ben-Neriah, “NF-kappaB Functions as a
Tumour Promoter in Inflammation-Associated Cancer,”
Nature, Vol. 431, No. 7007, 2004, pp. 461-466.
doi:10.1038/nature02924
[45] M. D. Brundage, D. Davies and W. J. Mackillop, “Prog-
nostic Factors in Non-Small Cell Lung Cancer: A Decade
of Progress,” Chest, Vol. 122, No. 3, 2002, pp. 1037-1057.
doi:10.1378/chest.122.3.1037
[46] S. Sugiura, Y. Ando, H. Minami, M. Ando, S. Sakai and
K. Shimokata, “Prognostic Value of Pleural Effusion in
Patients with Non-Small Cell Lung Cancer,” Clinical
Cancer Research, Vol. 3, 1997, pp. 47-50.
[47] N. Takigawa, Y. Segawa, M. Okahara, Y. Maeda, I. Ta-
kata, M. Kataoka and M. Fujii, “Prognostic Factors for
Patients with Advanced Non-Small Cell Lung Cancer:
Univariate and Multivariate Analyses Including Recur-
sive Partitioning and Amalgamation,” Lung Cancer, Vol.
15, No. 1, 1996, pp. 67-77.
doi:10.1016/0169-5002(96)00571-5
[48] K. S. Albain, J. J. Crowley, M. LeBlanc and R. B.
Livingston, “Survival Determinants in Extensive-Stage
Non-Small-Cell Lung Cancer: The Southwest Oncology
Group Experience,” Journal of Clinical Oncology, Vol. 9,
No. 9, 1991, pp. 1618-1626.
[49] A. Phillips, A. G. Shaper and P. H. Whincup, “Associa-
tion between Serum Albumin and Mortality from Car-
diovascular Disease, Cancer, and Other Causes,” Lancet,

Clinical Biomarkers and Prognosis in Taiwanese Patients with Non-Small Cell Lung Cancer (NSCLC) 423
Vol. 2, 1989, pp. 1434-1436.
doi:10.1016/S0140-6736(89)92042-4
[50] B. R. Don and G. Kaysen, “Serum Albumin: Relationship
to Inflammation and Nutrition,” Seminars in Dialysis, Vol.
17, No. 6, 2004, pp. 432-437.
doi:10.1111/j.0894-0959.2004.17603.x
[51] B. L. Sprague, A. Trentham-Dietz, B. E. Klein, R. Klein,
K. J. Cruickshanks, K. E. Lee and J. M. Hampton,
“Physical Activity, White Blood Cell Count, and Lung
Cancer Risk in a Prospective Cohort Study,” Cancer
Epidemiology, Biomarkers & Prevention, Vol. 17, No. 10,
2008, pp. 2714-2722.
doi:10.1158/1055-9965.EPI-08-0042
[52] Y. H. Li, S. H. Shieh and C. Y. Chen, “The Influence of
Health Behaviors on Survival in Lung Cancer Patients in
Taiwan,” Japanese Journal of Clinical Oncology, Vol. 41,
No. 3, 2011, pp. 365-372. doi:10.1093/jjco/hyq188
[53] T. A. Chiang, P. H. Chen, P. F. Wu, T. N. Wang, P. Y.
Chang, A. M. Ko, M. S. Huang and Y. C. Ko, “Important
Prognostic Factors for the Long-Term Survival of Lung
Cancer Subjects in Taiwan,” BMC Cancer, Vol. 8, No.
324, 2008.