j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

Available online at www.sciencedirect.com

ScienceDirect
journal homepage: www.JournalofSurgicalResearch.com

Prognostic value of ABO blood group in patients

with gastric cancer
Ye-Qiong Xu, MD,a,* Ting-Wang Jiang, PhD,b,* Yan-Hong Cui, MD,c
Yi-Lin Zhao, MD,c and Li-Qing Qiu, MDd
a

Molecular Laboratory, Changshu Medicine Examination Institute, Suzhou, Jiangsu, China

Department of Flow Cytometry, Changshu Medicine Examination Institute, Suzhou, Jiangsu, China
c
Molecular Laboratory, Changshu Medicine Examination Institute, Changshu, China
d
Laboratory Medicine, Dehua County Hospital, Quanzhou, China
b

article info

abstract

Article history:

Background: Previous studies have shown a link between the ABO blood groups and prog-

Received 8 June 2015

noses for several types of malignancies. However, little is known about the relationship

Received in revised form

between the ABO blood groups and prognosis in patients with gastric cancer (GC). The aim

26 October 2015

of this study was to investigate the prognostic performance of ABO blood groups in

Accepted 28 October 2015

patients with GC.

Available online 5 November 2015

Methods: A total of 1412 GC patients who had undergone curative intent surgery between
January 2005 and January 2010 participated in the present study. A prognostic nomogram

Keywords:

was constructed to improve the predictive capacity for patients with gastrectomy using R

ABO blood groups

software, and its predictive accuracy was determined by the concordance index (c-index).

Gastric cancer

Results: The median follow-up period of the 1412 GC patients was 44 mo with 809 alive.

Prognosis

Non-AB blood groups were associated with significantly decreased overall survival in GC
patients (hazard ratio 2.59; 95% confidence interval 1.74e3.86, P < 0.001), but patients in

Survival

the group AB had a better prognosis than those in the non-AB blood groups. Meanwhile,
group A had the worst prognosis among all the blood groups (hazard ratio 3.14;
95% confidence interval 2.09e4.72; P < 0.001). In addition, our constructed nomogram,
superior to that of the traditional AJCC stage system (c-index: 0.69), could more accurately
predict overall survival (c-index: 0.78) in GC patients who had undergone gastrectomy.
Conclusions: The blood group AB is a favorable prognostic factor for GC patients, but the
blood group A is an adverse prognostic factor for patients with gastrectomy. Further prospective studies are warranted to confirm this relationship.
2016 Elsevier Inc. All rights reserved.

1.

Introduction

Cancer presents a vital health problem for both health

care workers and patients. Cancer incidence rates vary
worldwide, and emerging evidence indicates that genetic
and environmental factors play a meaningful role in various

cancers. Gastric cancer (GC) is one of most common causes of

cancer death worldwide [1,2], and it is the second most common cancer and the third leading cause of death among
various types of malignancies in China. Although we have
achieved substantial progress in the diagnosis and treatment
of GC, the overall 5-y survival rate remains unsatisfactory due

* Corresponding authors. 36 Qingduntang Road, Changshu Medicine Examination Institute, Suzhou, Jiangsu, China. Tel./fax: 86
051252345208.
E-mail addresses: [email protected] (Y.-Q. Xu), [email protected] (T.-W. Jiang).
0022-4804/$ e see front matter 2016 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jss.2015.10.039

j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

to local relapse or metastasis after gastrectomy. Hence, the

survival rate will be obviously improved if GC can be detected
at an early tumor stage [3]. Better understanding of the association of genetic factors and predisposing environmental
factors of GC could improve patients prognoses and provide
appropriate therapy strategies.
The ABO blood group system is the first genetic polymorphism discovered in humans in 1990 [4]. Blood group
antigens are chemical components on the erythrocyte
membrane, and they are expressed in various epithelial
cells, including lung and gastrointestinal cells, urothelium,
mucosa, saliva, and body fluids [5]. The gene of the ABO blood
groups on chromosome 9q34 encodes glycosyltransferases that
catalyze the transfer of nucleotide donor sugars to H
antigens and form ABO blood group antigens [6]. The antigens
of the ABO blood groups are conveyed on the surface of red
blood cells and other tissues. The correlations of ABO blood
groups with benign or malignant cancers have also been
extensively studied for a long time. Aird et al. [7] reported that
ABO blood groups were closely associated with GC, and group A
was conspicuously more frequent in GC patients than normal
individuals.
Little is known about the connections between ABO blood
groups and clinical prognoses in patients with GC. Over the
past several decades, most studies have shed light on the
relationship between ABO blood groups and the risk of GC
[7e10]. However, the association of ABO blood groups with the
clinical prognosis of patients with gastrectomy remains
unclear. Herein, the aim of this study was to elucidate the
influence of ABO blood groups on the clinical outcomes of
patients with gastrectomy and construct a prognostic nomogram to improve the predictive capacity for overall survival
(OS) in patients with GC based on the ABO blood groups and
clinicopathologic parameters.

2.

Materials and methods

2.1.

Patients

This study was a retrospective analysis, and all newly diagnosed GC patients were obtained from the Changshu Medicine
Examination Institute and Nanjing First Hospital (Jiangsu,
China) between January 2005 and January 2010. The diagnosis
of GC was based on histologic evidence and classified according
to the seventh edition of the TNM-UICC/AJCC classification
system (Washington, 2010) [11]. The inclusion criteria were as
follows: (1) patients with gastric adenocarcinoma who had
undergone subtotal or total gastrectomy with full D2
lymphadenectomy; (2) all enrolled patients did not receive
preoperative anticancer therapy, such as radiotherapy or neoadjuvant chemotherapy; (3) the postoperation expected life
expectancy was 3 mo; (4) there was no infection or infectious
and hematologic diseases at preoperation; and (5) informed
consent was obtained from the eligible patients. Finally, 1412
patients were enrolled in this study. This study was approved
by the Medical Ethics Committee of Changshu Medicine
Examination Institute and Nanjing First Hospital, Nanjing
Medical University.

2.2.

189

Follow-up

Each patient was followed up periodically until death or

January 2015 (every 3 mo for the first 2 y, and every 6 mo up
to the fifth year) after surgery. Endoscopy, physical examination, and laboratory tests were conducted at each visit.
The follow-up cycles varied from 3e60 mo, with a median
of 44 mo. The OS ranged from the date of surgery to death.
The date of the last follow-up was applied to dropout
patients.

2.3.

Evaluation of ABO blood groups and covariates

Rh factors (positive or negative) and ABO blood groups (A, B,

AB, or O) were collected for complete medical histories.
Herein, we examined Rh factors and ABO blood groups in the
peripheral sera of enrolled participants. These determinations
were carried out at Changshu Medicine Examination Institute
and Nanjing First Hospital.
Among all the patients with GC, 717 (50.8%) received
adjuvant chemotherapy at postoperation, but 695 (49.2%) did
not receive any chemotherapy due to chemoresistance,
oppressive side effects, and other causes. Radiotherapy was
not administered in the patients routinely.

2.4.

Statistical analysis

Chi-square test was used to compare categorical variables.

The survival rates were evaluated by KaplaneMeier survival
analysis, and their significance was calculated by the log-rank
test. The predictors for OS tested by univariate analysis were
calculated by multivariate analysis using the Cox proportional
hazards model. The nomogram for OS was established using
R 3.0.3 software (Institute for Statistics and Mathematics,
Vienna, Austria). The predictive accuracy was calculated
by the Harrells concordance index (c-index). All the statistical
analyses were conducted using IBM SPSS 20.0 software (IBM,
Chicago, IL). A P value <0.05 was considered to be statistically
significant.

3.

Results

3.1.

Clinicopathologic characteristics of the GC patients

There were 1412 enrolled GC patients, 968 (68.6%) male and

444 (31.4%) female, aged from 29e87 y (the median age was
64 y). Based on the seventh edition of the TNM-UICC/AJCC
classification system, the numbers of patients with stage I,
II, and III disease were 339 (24.0%), 404 (28.6%), and 669 (47.4%),
respectively. The numbers of GC patients with tumor grade G1
and G2eG3 were 282 (20.0%) and 1130 (80.0%), respectively. In
addition, the cutoff values of serum CA199 and CA724 were 37
and 6 U/mL, respectively. The ABO blood groups were
distributed as 143 (10.1%) with blood group AB and 1269
(89.9%) with non-AB blood groups (407 patients with the blood
group O, 346 patients with the blood group B, and 516 patients
with the blood group A), and the characteristics of the enrolled
participants are summarized in Table 1. Our results indicated
that patients with the blood group AB were obviously

190

j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

Table 1 e Baseline patient characteristics according to ABO blood group.

Characteristic
total
Rh
Positive
Negative
Sex
Male
Female
Age
64
>64
Hp
No
Yes
Lauren
Intestinal type
Mixed type
Diffuse type
Grade
G1
G2eG3
Stagey

Depth of invasion
T1
T2
T3eT4
Lymph node
N0
N1eN3
Chemotherapy
None
Adjuvant
CA199 (U/mL)
37
>37
CA724 (U/mL)
6
>6

P*

Number of patients
(%) 1412

O (%) 407

A (%) 516

B (%) 346

AB (%) 143

Non-AB (%) 1269

1404 (99.4)
8 (0.6)

405 (99.5)
2 (0.5)

142 (99.6)
2 (0.4)

373 (99.1)
3 (0.9)

142 (99.3)
1 (0.7)

1262 (99.4)
7 (0.6)

0.715

968 (68.6)
444 (31.4)

314 (77.1)
93 (22.9)

333 (64.5)
183 (35.5)

236 (68.2)
110 (31.8)

85 (59.4)
58 (40.6)

883 (69.6)
386 (30.4)

0.013

775 (54.9)
637 (45.1)

247 (60.7)
202 (39.3)

280 (54.3)
236 (45.7)

143 (41.3)
203 (58.7)

105 (73.4)
38 (26.6)

670 (52.8)
599 (47.2)

<0.001

300 (21.2)
1112 (78.8)

88 (21.6)
319 (78.4)

93 (18.0)
423 (82.0)

99 (28.6)
247 (71.4)

20 (18.8)
123 (81.2)

280 (22.1)
989 (77.9)

0.025

453 (32.1)
326 (23.1)
633 (44.8)

131 (32.2)
101 (24.8)
175 (43.0)

168 (32.6)
102 (19.8)
246 (47.7)

101 (29.2)
94 (27.2)
151 (43.6)

53 (37.1)
29 (20.3)
61 (42.7)

400 (31.5)
297 (23.4)
572 (45.1)

0.379

282 (20.0)
1130 (80.0)

83 (20.4)
324 (79.6)

90 (17.4)
426 (82.6)

71 (20.5)
275 (79.5)

38 (26.6)
105 (73.4)

244 (19.2)
1025 (80.8)

0.037

339 (24.0)
404 (28.6)
669 (47.4)

97 (23.8)
119 (29.2)
191 (46.9)

123 (23.8)
150 (29.1)
243 (47.1)

84 (24.3)
97 (28.0)
165 (47.7)

35 (24.5)
38 (26.6)
70 (49.0)

304 (24.0)
366 (28.8)
599 (47.2)

0.849

216 (15.3)
372 (26.3)
824 (58.4)

68 (16.7)
107 (26.3)
232 (57.0)

72 (14.0)
139 (26.9)
305 (59.1)

53 (15.3)
82 (23.7)
211 (61.0)

23 (16.1)
44 (30.8)
76 (53.1)

193 (15.2)
328 (25.8)
748 (58.9)

0.370

513 (36.3)
899 (63.7)

152 (37.3)
255 (62.7)

196 (38.0)
320 (62.0)

121 (35.0)
225 (65.0)

44 (30.8)
99 (69.2)

469 (37.0)
800 (63.0)

0.145

695 (49.2)
717 (50.8)

185 (45.5)
222 (54.5)

259 (50.2)
257 (49.8)

196 (56.6)
150 (43.4)

55 (38.5)
88 (61.5)

640 (50.4)
629 (49.6)

0.007

766 (54.2)
646 (45.8)

214 (52.6)
193 (47.4)

281 (54.5)
235 (45.5)

186 (53.8)
160 (46.2)

85 (59.4)
58 (40.6)

681 (53.7)
588 (46.3)

0.189

801 (56.7)
611 (43.3)

233 (57.2)
174 (42.8)

286 (55.4)
230 (44.6)

187 (54.0)
159 (46.0)

95 (66.4)
48 (33.6)

706 (55.6)
563 (44.4)

0.013

Hp Helicobacter pylori.
*
Differences between the blood group AB and non-AB blood groups were tested by chi-square test.
y
Tumor stage according to the seventh edition of the American Joint of Committee on Cancer.

associated with sex (P 0.013), age (P < 0.001), Helicobacter

pylori infection (P 0.025), grade (P 0.037), chemotherapy
(P 0.007), and CA724 (P 0.013).

3.2.
Association between the clinical prognosis of GC
patients and ABO blood groups
The median follow-up period of the 1412 GC patients was
44 mo with 809 (57.3%) alive and 603 (42.7%) dead from
cancer-related diseases at the final clinical follow-up. The
ABO blood groups were closely associated with OS according to the KaplaneMeier analysis. OS was longer in GC
patients with the group AB (median, 58 mo) than in those
with the group O (median, 43 mo), group B (median, 38 mo),
and group A (median, 30 mo; P < 0.001) as shown in
Figure 1A. Furthermore, similar results were observed when

stratified by tumor stage (Fig. 1BeD; P < 0.001). In addition,

we evaluated the influence of the patients Rh, ABO blood
groups, sex, age, H pylori status, Lauren classification, grade,
stage, depth of invasion, lymph node, chemotherapy,
CA199, and CA724 on OS by univariate analysis (Table 2).
The OS was significantly associated with the patients sex,
age, Lauren classification, grade, tumor stage, depth of invasion, lymph node, chemotherapy, CA199, CA724, and ABO
blood groups (P < 0.05). Meanwhile, the significant parameters in univariate analysis were further analyzed in
multivariate analyses. The results showed that the ABO
blood groups were independent predictors for OS (group A:
hazard ratio [HR] 3.14, 95% confidence interval
[CI] 2.09e4.72; group B: HR 2.52, 95% CI 1.65e3.83;
group O: HR 2.06, 95% CI 1.35e3.15; and non-AB blood
groups: HR 2.59, 95% CI 1.74e3.86; P < 0.001; Table 2).

j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

191

Fig. 1 e KaplaneMeier survival curves for OS based on ABO blood type. (A) OS for the entire cohort of patients with gastric
cancer; (B) OS for tumor stage I patients with gastric cancer; (C) OS for tumor stage II patients with gastric cancer; and (D) OS
for tumor stage III patients with gastric cancer. (Color version of the figure is available online.)

3.3.

Prognostic nomogram for OS

A prognostic nomogram was developed to predict the survival

of GC patients after surgical resection using the seven independent factors identified in the multivariate analysis
(Table 2). The nomogram was used by summing the points
identified on the top scale for each independent factor. This
total point score was then identified on the total points scale
to determine the probability of 3- and 5-y survival (Fig. 2).
The c-index for this nomogram was 0.78 based on the fitted
multivariable Cox model. The calibration curve (Fig. 3) showed
how the predictions from the nomogram compared with
actual outcomes for the 1412 patients. The dashed line

represented the performance of an ideal nomogram, in which

the predicted consequences perfectly matched with the actual
consequences. In addition, traditional AJCC stage groups and
nomogram predictions were compared for their ability to
distinguish among the patients by the c-index. Our constructed nomogram was superior to that of traditional AJCC
stage groups (c-index: 0.78 versus 0.69; P < 0.001).

4.

Discussion

The antigens of the ABO blood groups have been investigated in

the development of malignancy in the past several decades.

192

j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

Table 2 e Univariate and multivariate analysis of the prognostic factors of OS by the Cox regression model.
Variables

Rh
Positive
Negative
Sex
Male
Female
Age
64
>64
Hp
No
Yes
Lauren
Intestinal type
Mix type
Diffuse type
Grade
G1
G2eG3
Stage*

Depth of invasion
T1
T2
T3eT4
Lymph node
N0
N1eN3
Chemotherapy
None
Adjuvant
CA199 (U/mL)
37
>37
CA724 (U/mL)
6
>6
Blood group
AB
A
B
O
Non-AB

Univariate analysis
HR

95% CI

1.00
0.23

Reference
0.03e1.64

1.00
1.19

Multivariate analysis
P

HR

95% CI

0.143

d
d

d
d

d
d

Reference
1.01e1.41

0.042

1.00
1.15

Reference
0.97e1.36

0.119

1.00
1.37

Reference
1.17e1.61

<0.001

1.00
1.32

Reference
1.12e1.55

0.001

1.00
1.17

Reference
0.96e1.44

0.126

d
d

d
d

1.00
2.26
3.36

Reference
1.75e2.92
2.69e4.19

<0.001
<0.001

1.00
1.00
1.46

Reference
0.41e2.43
0.61e3.49

0.994
0.395

1.00
4.93

Reference
3.53e6.88

<0.001

1.00
2.96

Reference
2.11e4.15

<0.001

1.00
4.52
8.98

Reference
3.12e6.56
6.33e12.74

<0.001
<0.001

1.00
1.55
1.81

Reference
0.95e2.51
1.03e3.18

0.078
0.040

1.00
2.15
6.43

Reference
1.39e3.32
4.33e9.55

0.001
<0.001

1.00
2.72
4.56

Reference
1.26e5.88
2.87e7.23

0.011
<0.001

1.00
3.11

Reference
2.53e3.81

<0.001

1.00
2.53

Reference
2.05e3.12

<0.001

1.00
0.65

Reference
0.56e0.77

<0.001

1.00
0.71

Reference
0.60e0.84

<0.001

1.00
1.20

Reference
1.03e1.41

0.023

1.00
0.95

Reference
0.67e1.28

0.216

1.00
2.07

Reference
1.76e2.44

<0.001

1.00
2.07

Reference
1.67e2.55

<0.001

1.00
3.73
2.96
2.17
3.00

Reference
2.49e5.57
1.95e4.48
1.43e3.30
2.03e4.45

<0.001
<0.001
<0.001
<0.001

1.00
3.14
2.52
2.06
2.59

Reference
2.09e4.72
1.65e3.83
1.35e3.15
1.74e3.86

<0.001
<0.001
0.001
<0.001

d
d

Hp Helicobacter pylori.
*
Tumor stage according to the seventh edition of American Joint of Committee on Cancer.

Previous research has found an obvious relationship between

the incidence of malignancies and ABO blood groups [12];
however, many investigations have shown different associations between ABO blood groups and GC patients [13,14].
However, few studies have explored the prognostic value of
ABO blood groups in surgical GC patients. Hence, our study was
aimed to investigate the impact of ABO blood groups on the
clinical prognosis of 1412 GC patients who had undergone
gastrectomy and first attempted to establish a predictive model
to improve the predictive accuracy in patients with surgical GC.
A total of 1412 GC patients were enrolled in the retrospective study. Our results indicated that non-AB blood

groups were associated with significantly decreased OS in GC

patients, but patients with group AB had a better prognosis
than those with non-AB blood groups. Meanwhile, group A
had the worst prognosis among all the blood groups
(HR 3.14; 95% CI 2.09e4.72; P < 0.001). However, Xu et al.
[13] reported that there was no significant difference between
ABO blood groups and the survival of 474 GC patients, a
finding that was inconsistent with our outcome. These contradictory consequences may be explained as follows. First,
our enrolled patients had undergone gastrectomy with full D2
lymphadenectomy. Second, as patients with stage disease
had experienced local relapse or distant metastasis, many

j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

193

Fig. 2 e Nomogram with significant clinical parameters predicts the probability of postoperative patients with
adenocarcinoma for the 3- or 5-y overall survival.

patients typically undergo various heterogeneous treatments.

Herein, patients with stage disease were not evaluated in
the present study. Third, different numbers of samples could
induce a different impact of the ABO blood groups on prognosis of GC patients. However, in the present study, there was
the higher prevalence of younger patients in the AB blood
group than the non-AB blood groups (median age: 61 versus 64
y). To explore the bias, we also investigated the influence of
blood group AB on GC development in 327 patients <61 y. We
found that group AB was susceptible to better prognosis than
those with non-AB blood groups (P 0.042), and group A
significantly reduced survival rates of GC patients (P 0.001).
Moreover, ABO blood type was an independent prognostic
factor from the multivariate analysis. Therefore, the bias did

Fig. 3 e Calibration plots for the 5-y overall survival. The

x-axis is the nomograms predicted probability of overall
survival, and the y-axis is the actual probability of overall
survival estimated by the KaplaneMeier method. The
dashed line represents an ideal nomogram, and the solid
line indicates performance of the present nomogram.

not affect our results. In addition, this study revealed a

significant link between ABO blood groups and H pylori infection, a result that was similar to that in a previous study in GC
[15]; however, no significant association between H pylori and
survival of GC patients was observed in our study. Meanwhile,
adjuvant chemotherapy was closely associated with
increased OS compared with nonchemotherapy. This may be
due to depressed tumor growth and tumor cell death after
surgery to reduce local relapse and distant metastasis by
adjuvant chemotherapy.
Previous studies have observed relationships between the
clinical prognosis and ABO blood groups, but their results
were inconsistent. Research from Ouyang et al. [16] has
reported that group A was associated with poor prognosis in
nasopharyngeal carcinoma. By contrast, there was no significant association between nasopharyngeal carcinoma and
ABO blood groups [17], and similar results occurred in colon
cancer [18,19]. Some potential mechanisms may support the
association between cancer and ABO blood groups, including
inflammation, immunosurveillance of the tumor cell lines,
membrane signaling, and intracellular adhesion [20].
ABO blood groups are named according to the carbohydrate moieties shown on the surface of red blood cells and
expressed on membrane lipids and proteins. The variant
alleles (O, B, and A) of a single gene, located on the chromosomes 9q34, determine human blood groups by encoding
three glycosyltransferases with diverse substrate specificities.
The antigens of the ABO blood groups are overexpressed on
the surface of epithelial cells, such as the skin, urogenital,
bronchopulmonary, and gastrointestinal cells [21,22]. The
glycosyltransferase specificity has a significant impact beyond
naming ABO blood groups. The glycoconjugates are important
mediators of membrane signaling and intercellular adhesion
in two processes integral to tumor spread and progression
[22]. Moreover, surface molecules, distinguished by the host
immune response, play a role in facilitating immunosurveillance for the tumor cells [21].
Alterations in the host inflammatory response as a result
of ABO blood group antigens can offer a further mechanism to
explain the relationships between GC and ABO blood groups.
The relationship between malignancies and inflammation

194

j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

was first reported in 1863, explaining that the lymphoreticular infiltrate reflected original cancer at the sites of
chronic inflammation [23]. The influence of the inflammatory
microenvironment on malignancy has been supported by
emerging evidence over the past decades. Deeper understanding of the relationships between inflammation and
malignancies contributed to the treatment and prevention of
malignancies. These previous results have suggested that
chronic inflammation is closely related to tumor metastasis
and initiation and revealed that ABO blood groups could
impact the clinical prognosis in patients with cancer.
Nomograms have been developed in different cancers and
have been shown to be more accurate than conventional
staging systems for predicting prognosis in various cancers
[24e26]. Furthermore, nomograms have been validated to be
more accurate for predicting the disease-specific survival rate
of GC in Western countries by combining all the independent
prognostic factors and quantified risks [27e30]. A previous
study externally validated the Memorial Sloan Kettering
Cancer Center nomogram in predicting the probability of 5and 9-y disease-specific survival after R0 resection for GC. The
Memorial Sloan Kettering Cancer Center nomogram performed well with good discrimination and calibration [31].
Therefore, the present study attempted to establish a predictive nomogram to predict the probability of postoperative
patients who will die of cancer-related causes within 3 and 5 y
based on ABO blood groups and independent factors in
multivariate analyses. The nomogram performed well in
predicting OS, and the prediction was supported by the
c-index (0.78), a finding that was superior to that of the
traditional AJCC stage system (c-index: 0.69). All the results
supported that nomograms may forecast well the clinical
outcome in patients with GC at postoperation.
Some limitations of the retrospective study should be
acknowledged. First, in this study, some patients did not
undergo computed tomography or magnetic resonance
imaging of the chest and/or brain at diagnosis, and it is
possible that many patients had asymptomatic disease at the
time of primary treatment. Second, all the participants of our
study were Chinese, perhaps limiting the generalizability of
this result. Further investigations are urgently needed to
clarify our results, including the evaluation of different races
from other places. Third, there was a selection bias of younger
patients more likely to receive adjuvant chemotherapy than
patients >64 y in the AB versus non-AB groups; however, the
multivariate analysis showed that ABO blood groups were
independent predictors for OS, particularly independent of
age and receipt of chemotherapy.

5.

Conclusions

In conclusion, our data suggested an important link between

surgical GC survival and ABO blood groups. GC patients with
the group AB were more susceptible to show promising
prognostic significance than those with non-AB blood groups
(groups O, A, and B). GC patients with the blood group A had a
worse survival rate than those with the blood groups O, B, and
AB. Our constructed nomogram did well to predict the clinical
prognosis in patients with surgical GC.

Acknowledgment
This study was supported by the Science and Technology
Development Project of Changshu (201218, 201313, 201417).
Authors contributions: Y.-Q.X. conceived and designed the
experiments and wrote the article. Y.-Q.X., T.-W.J., and L.-Q.Q.
performed the experiments. Y.-H.C. analyzed the data. Y.-L.Z.
contributed to the reagents/materials/analysis tools. Y.-Q.X.
and T.-W.J. designed the software used in analysis.

Disclosure
The authors declare that they have no competing interest.

references

[1] Ferlay J, Shin HR, Bray F, et al. Estimates of worldwide burden

of cancer in 2008: GLOBOCAN 2008. Int J Cancer 2010;127:
2893.
[2] Shah MA, Kelsen DP. Gastric cancer: a primer on the
epidemiology and biology of the disease and an overview of
the medical management of advanced disease. J Natl Compr
Canc Netw 2010;8:437.
[3] Yoo CH, Noh SH, Shin DW, Choi SH, Min JS. Recurrence
following curative resection for gastric carcinoma. Br J Surg
2000;87:236.
[4] Yamamoto F, Cid E, Yamamoto M, Blancher A. ABO research
in the modern era of genomics. Transfus Med Rev 2012;26:103.
[5] Graziano SL, Tatum AH, Gonchoroff NJ, Newman NB,
Kohman LJ. Blood group antigen A and flow cytometric
analysis in resected early-stage non-small cell lung cancer.
Clin Cancer Res 1997;3:87.
[6] Yazer MH. What a difference 2 nucleotides make: a short
review of ABO genetics. Transfus Med Rev 2005;19:200.
[7] Aird I, Bentall HH, Roberts JA. A relationship between cancer
of stomach and the ABO blood groups. Br Med J 1953;1:799.
[8] Beasley WH. Blood groups of gastric ulcer and carcinoma. Br
Med J 1960;1:1167.
[9] Roberts JA. Some associations between blood groups and
disease. Br Med Bull 1959;15:129.
[10] Wiener AS. Blood-groups and disease. A critical review.
Lancet 1962;1:813.
[11] Washington K. 7th edition of the AJCC cancer staging
manual: stomach. Ann Surg Oncol 2010;17:3077.
[12] Hoskins LC, Loux HA, Britten A, Zamcheck N. Distribution of
ABO blood groups in patients with pernicious anemia, gastric
carcinoma and gastric carcinoma associated with pernicious
anemia. N Engl J Med 1965;273:633.
[13] Qiu MZ, Zhang DS, Ruan DY, et al. A relationship between
ABO blood groups and clinicopathologic characteristics of
patients with gastric adenocarcinoma in China. Med Oncol
2011;28(Suppl 1):S268.
[14] Li B, Tan B, Chen C, Zhao L, Qin L. Association between the
ABO blood group and risk of common cancers. J Evid Based
Med 2014;7:79.
[15] Nakao M, Matsuo K, Ito H, et al. ABO genotype and the risk of
gastric cancer, atrophic gastritis, and Helicobacter pylori
infection. Cancer Epidemiol Biomarkers Prev 2011;20:1665.
[16] Ouyang PY, Su Z, Mao YP, Liu Q, Xie FY. Prognostic value
of ABO blood group in southern Chinese patients with
established nasopharyngeal carcinoma. Br J Cancer 2013;
109:2462.

j o u r n a l o f s u r g i c a l r e s e a r c h 2 0 1 ( 2 0 1 6 ) 1 8 8 e1 9 5

[17] Zhang YX, Kang SY, Chen G, et al. ABO blood group, EpsteinBarr virus infection and prognosis of patients with nonmetastatic nasopharyngeal carcinoma. Asian Pac J Cancer
Prev 2014;15:7459.
[18] Cao X, Wen ZS, Sun YJ, et al. Prognostic value of ABO blood
group in patients with surgically resected colon cancer. Br J
Cancer 2014;111:174.
[19] Slater G, Itzkowitz S, Azar S, Aufses AH Jr. Clinicopathologic
correlations of ABO and Rhesus blood type in colorectal
cancer. Dis Colon Rectum 1993;36:5.
[20] Nozoe T, Ezaki T, Baba H, Kakeji Y, Maehara Y. Correlation of
ABO blood group with clinicopathologic characteristics of
patients with esophageal squamous cell carcinoma. Dis
Esophagus 2004;17:146.
[21] Le Pendu J, Marionneau S, Cailleau-Thomas A, et al. ABH and
Lewis histo-blood group antigens in cancer. APMIS 2001;109:9.
[22] Hakomori S. Antigen structure and genetic basis of histoblood groups A, B and O: their changes associated with
human cancer. Biochim Biophys Acta 1999;1473:247.
[23] Balkwill F, Mantovani A. Inflammation and cancer: back to
Virchow? Lancet 2001;357:539.
[24] Sternberg CN. Are nomograms better than currently
available stage groupings for bladder cancer? J Clin Oncol
2006;24:3819.

195

[25] Touijer K, Scardino PT. Nomograms for staging, prognosis,

and predicting treatment outcomes. Cancer 2009;115:3107.
[26] Deng Q, He B, Gao T, et al. Up-regulation of 91H promotes
tumor metastasis and predicts poor prognosis for patients
with colorectal cancer. PLoS One 2014;9:e103022.
[27] Kattan MW, Karpeh MS, Mazumdar M, Brennan MF.
Postoperative nomogram for disease-specific survival
after an R0 resection for gastric carcinoma. J Clin Oncol 2003;
21:3647.
[28] Gold JS, Tang LH, Gonen M, et al. Utility of a prognostic
nomogram designed for gastric cancer in predicting outcome
of patients with R0 resected duodenal adenocarcinoma. Ann
Surg Oncol 2007;14:3159.
[29] Peeters KC, Kattan MW, Hartgrink HH, et al. Validation of a
nomogram for predicting disease-specific survival after an
R0 resection for gastric carcinoma. Cancer 2005;103:702.
[30] Novotny AR, Schuhmacher C, Busch R, et al. Predicting
individual survival after gastric cancer resection: validation
of a U.S.-derived nomogram at a single high-volume center
in Europe. Ann Surg 2006;243:74.
[31] Chen D, Jiang B, Xing J, et al. Validation of the memorial
Sloan-Kettering Cancer Center nomogram to predict diseasespecific survival after R0 resection in a Chinese gastric
cancer population. PLoS One 2013;8:e76041.