Keywords
young gastric cancer - aggressive histology - gastric adenocarcinoma
Introduction
Gastric cancer (GC) is a very aggressive disease and remains one of the most common
causes of cancer-related mortality worldwide. The age-standardized incidence rate
for gastric cancer is 4.5 per 1 lakh population. Two-thirds of the patients present
at an advanced stage due to non-specific symptoms at presentation, finally ending
in palliative treatment.[1]
Across the literature, variation exists in the specific cut-offs used to define gastric
cancer in young (GCY) adult patients. Most studies used an age cut-off of less than
40 years for GCY. Our study described GCY as all diagnosed gastric adenocarcinoma
patients aged up to 30 years. Gastric cancer in older adults (GCO) includes all GC
patients aged above 30 years. Our study's median age of GC is one decade less than
that seen in western countries, so in our research, we defined GCY as less than 30
years.[2]
Various studies have reported the incidence of GCY between 2 and 8%.[3]
[4]
[5]
[6] The age-adjusted incidence rate in GCY from the SEER database was 0.9 per 1 lakh.
The incidence of GC has decreased worldwide, but the incidence of GCY has increased
over the past decade.[6] A single-center study from India reported the incidence of GCY as 18%, which was
high compared to the literature data.[7]
GCY faces challenges such as biological tumor variation, advanced-stage diagnosis,
treatment adherence, fertility preservation problems, and psychosocial considerations.[4]
[6]
[7] This study aimed to find the differences in clinical characteristics and treatment
outcomes of GCY compared to GCO.
Materials and Methods
Study Population
This is a retrospective study in a tertiary care center in Southern India. We screened
records from 2015 to 2020 and identified 33 records of GCY. To provide comparative
data, we used the retrospective data of GCO from 2015 to 2018, which was available.
Of the 505 patients, who presented with GC to our department from January 2015 to
December 2018, 469 patients were >30 GCO with adenocarcinoma histology. The consort
diagram of inclusion criteria and age is represented in [Fig. 1].
Fig. 1 Consort flow diagram of study participants.
Statistical Analysis
Categorical data are expressed in proportion, and a median described the continuous
data with a range. The association between GCY and GCO patients was studied using
a chi-square test. Overall survival (OS) was defined as the time from the date of
diagnosis to death. For the patients alive at the last follow-up, OS was censored
at the last follow-up or April 30, 2021, whichever came first. The Kaplan–Meier method
was used to estimate the survival curves, and the log-rank test was used to compare
survival data. IBM SPSS ver.19 was used for the analysis of the data. A p-value < 0.05 was considered statistically significant.
Results
From our records, we identified 33 patients with GCY from 2015 to 2020. The median
age was 28 years (21–30), with a male to female ratio of 1:2. The most common symptom
at presentation was abdomen pain, followed by weight loss and vomiting. Gastric outlet
obstruction at presentation was seen in 22%. [Table 1] describes the baseline characteristics. Distal GC were seen more often than proximal
cancers. In GCY, 60% of patients presented with metastatic disease and 40% with nonmetastatic
disease. Among the 40% of patients with non-metastatic disease, most (61%) were locally
advanced (T4a and T4b). Nearly all patients had nodal involvement. In patients with
metastasis, peritoneum, liver, and nonregional nodes were the most common sites of
metastasis. Tumor characteristics of all patients are reported in [Table 2].
Table 1
Baseline comparison of clinical and demography characteristics of GCY and GCO
|
S. No.
|
Variable
|
Category
|
Gastric cancer in young (GCY)
(n = 33)
|
Gastric cancer in old (GCO)
(n = 469)
|
p-Value
|
|
1.
|
Age (y)
|
Median (range)
|
28 (21-30)
|
55 (31-86)
|
|
|
2.
|
Gender
|
Female
|
22 (66.7)
|
148 (31.6)
|
<0.001
|
|
Male
|
11 (33.3)
|
321 (68.4)
|
|
3.
|
BMI
|
Underweight
|
23 (74.2)
|
207 (56.1)
|
0.06
|
|
Non-underweight
|
8 (25.8)
|
162 (43.9)
|
|
Missing data
|
2
|
100
|
|
4.
|
ECOG
|
0-1
|
30 (90.9)
|
322 (68.7)
|
0.02
|
|
2
|
3 (9.1)
|
124 (26.4)
|
|
3-4
|
|
23 (4.9)
|
|
5.
|
Co-morbidity
|
Yes
|
2 (6.1)
|
80 (17.1)
|
0.09
|
|
6.
|
Type of comorbidities
|
Diabetes Mellitus
|
0 (0.0)
|
32(40)
|
|
|
Hypertension
|
0 (0.0)
|
18 (22.5)
|
|
|
DM and Hypertension
|
0 (0.0)
|
12 (15)
|
|
|
Chronic Obstructive Pulmonary Disease
|
0 (0.0)
|
5 (6.3)
|
|
|
others
|
2 (100)
|
13 (16.2)
|
|
|
7.
|
Albumin
|
Median (range)
|
4.1 (2.3-4.7)
|
3.5 (1.2-4.9)
|
0.002
|
|
8.
|
Duration of symptoms
(months)
|
Median (range)
|
2 (0.10-12)
|
3 (0.1–24)
|
0.32
|
|
9.
|
Type of symptoms
|
Abdominal pain
|
23 (69.7)
|
301 (64.2)
|
|
|
Vomiting
|
25 (75.8)
|
269 (57.4)
|
|
|
Loss of weight
|
21 (63.6)
|
262 (55.9)
|
|
|
Loss of appetite
|
23 (69.7)
|
213 (45.4)
|
|
|
Dyspepsia
|
11 (33.3)
|
47 (10)
|
|
|
Melena
|
6 (18.2)
|
81 (17.3)
|
|
|
Abdominal distension
|
2(6.1)
|
43(9.2)
|
|
|
Hematemesis
|
1 (3.0)
|
30 (6.4)
|
|
|
Mass abdomen
|
1(3.0)
|
21(4.5)
|
|
|
10.
|
Gastric outlet obstruction
|
Positive
|
7 (21.2)
|
101 (40.2)
|
0.04
|
Abbreviations: BMI, body mass index; ECOG, Eastern cooperative group.
Table 2
Baseline comparison of tumor characteristics of GCY and GCO
|
S. No.
|
Variable
|
Category
|
GCY (n = 33)
|
GCO (n = 469)
|
p-Value
|
|
1.
|
Site of tumor
|
GEJ/cardia
|
9 (27.3)
|
97 (22.2)
|
0.53
|
|
Fundus and body
|
6 (18.2)
|
60 (13.8)
|
|
Antrum and Pylorus
|
18 (54.5)
|
262 (60.1)
|
|
Linitis plastica
|
0 (0.0)
|
17 (3.9)
|
|
2.
|
Stage
|
Early
|
1 (3.0)
|
6 (1.3)
|
0.43
|
|
Locally advanced
|
12 (36.4)
|
216 (46.1)
|
|
Metastasis
|
20 (60.6)
|
247 (52.7)
|
|
3.
|
TNM staging (early and locally advanced)
|
|
|
|
|
|
T status
|
T1-3
|
5 (38.5)
|
71 (32.9)
|
0.41
|
|
T4a
|
6 (46.2)
|
74 (34.3)
|
|
T4b
|
2 (15.4)
|
71 (32.9)
|
|
N status
|
N0-N1
|
9 (69.2)
|
131 (61.2)
|
0.77
|
|
N2-N3
|
4 (30.8)
|
83 (38.8)
|
|
Missing
|
|
2
|
|
|
4.
|
Metastasis
|
Single
|
11(55.0)
|
207 (84.8)
|
0.003
|
|
Multiple
|
9[a] (45.0)
|
37 (15.2)[b]
|
|
Missing data
|
|
3
|
|
|
5.
|
Site of single metastasis
|
Organ metastates
|
9 (81.8)
|
155 (74.9)
|
|
|
Liver
|
1 (9.1)
|
65 (31.4)
|
|
|
bone
|
0 (0.0)
|
3 (1.4)
|
|
|
lung
|
0 (0.0)
|
5 (2.4)
|
|
|
Omental
|
2 (18.2)
|
0 (0.0)
|
|
|
Adrenal
|
0 (0.0)
|
4 (1.9)
|
|
|
Peritoneal
|
2 (18.2)
|
69 (33.3)
|
|
|
Ovary
|
4 (36.4)
|
7 (3.4)
|
|
|
Kidney
|
0 (0.0)
|
1 (0.5)
|
|
|
Skin
|
0
|
1(0.5)
|
|
|
Nonregional nodal metastases
|
2(18.2)
|
52(25.1)
|
|
|
6.
|
Histopathology subtype
|
Diffuse
|
22 (66.7)
|
186 (41.7)
|
0.001
|
|
a. Signet
|
10 (45.5)
|
65 (14.6)
|
|
b. Nonsignet
|
12 (54.5)
|
121 (27.1)
|
|
Intestinal
|
8 (24.2)
|
259 (58.1)
|
|
Mixed
|
|
1 (0.2)
|
|
Missing data
|
3
|
23
|
|
|
7.
|
Treatment
|
|
|
|
|
|
Surgery
|
Curative
|
6 (18.2)
|
102 (28.3)
|
0.17
|
|
Palliative
|
7 (21.2)
|
118 (32.8)
|
|
Chemotherapy
|
NACT
|
5 (15.2)
|
62 (13.2)
|
0.46
|
|
Adjuvant
|
3 (9.1)
|
63 (13.4)
|
|
Palliative
|
22 (66.7)
|
260 (55.4)
|
|
No chemo
|
3 (9.1)
|
84 (17.9)
|
a
GCY = Site of multiple metastasis (9):1 (liver, bone); 2 (liver, peritoneal); 1 (peritoneal,
kidney); 2 (peritoneal, mediastinal); 2 (nodal, peritoneal) and 1 (ovary, nodal).
b
GCO = Site of multiple metastasis (37): 4 (liver, nodal); 1 (liver, nodal, and bone);
3 (liver, nodal, peritoneal) ; 1 (liver, nodal, and ovary), 1 (liver and bone); 1
(liver, bone, lungs); 7 (liver and peritoneal); 3 (liver and lungs); 9 (nodal and
peritoneal); 1 (nodal, lungs); 2 (bone and peritoneal); 3 (peritoneal and ovary);
1 (bone and adrenal).
A comparison of the clinical and pathological parameters between GCY and GCO is shown
in [Tables 1] and [2]. Males were predominant in GCO (68%), whereas in the GCY group, females were predominant
(67%). No patients in GCY had a family history of GC. GCY had good performance status
(PS 1) compared to GCO (90.9% vs. 68.7%, p-value 0.02). Both groups had distal gastric cancers more often than proximal cancers,
and GCY had less incidence of gastric outlet obstruction than GCO (21% vs. 40%, p-value 0.04). Diffuse-type tumor histology was seen more in the GCY than in GCO (66.7%
vs. 41.7%, p-value 0.001). In patients with metastasis, multiple metastases were common in GCY
compared to GCO (45% vs. 15%, p-value 0.003).
Of the 33 patients in GCY, 8 (24%) patients received curative-intent treatment (including
perioperative chemotherapy with surgery), 22 (67%) received palliative chemotherapy,
and 3 (9%) received best supportive care. The commonest chemotherapy regimen used
in the curative setting was the EOX regimen in 38%. Among those patients treated with
curative intent in GCY, dose modification was done in two patients (25%) due to chemotoxicity
in the previous cycle, and delay in the chemotherapy was recorded in five patients
(75%). The commonest chemotherapy regimen used in the palliative setting was the EOX
regimen in 59% of GCY. Of the 22 patients who received first-line palliative chemotherapy
on progression, only three were fit to receive second-line chemotherapy in the GCY
group. The reasons for dose modification and delay in chemotherapy in GCY are detailed
in [Table 3].
Table 3
Gastric cancer treatment modality in GCY and GCO
|
S.
No.
|
Chemotherapy
|
Subcategory
|
Gastric cancer in young (GCY)
(n = 30)
|
Gastric cancer in old (GCO)
(n = 385)
|
|
NACT (N = 5)
|
Adjuvant
(N = 3)
|
Palliative
(N = 22)
|
NACT (N = 62)
|
Adjuvant
(N = 63)
|
Palliative
(N = 260)
|
|
|
1
|
First line
|
FLOT
|
2 (40)
|
|
3 (13.6)
|
|
|
|
|
|
EOX
|
3 (60)
|
|
13 (59.2)
|
55 (88.7)
|
15 (23.8)
|
183 (70.3)
|
|
|
CAP-CIS
|
|
1 (33.3)
|
|
|
16 (25.4)
|
1 (0.4)
|
|
|
CAPOX
|
|
1 (33.3)
|
2 (9.1)
|
2 (3.2)
|
15 (23.8)
|
24 (9.2)
|
|
|
Capecitabine
|
|
|
3 (13.6)
|
5 (8.1)
|
13 (20.6)
|
47 (18.1)
|
|
|
others
|
|
1 (33.3)
|
1(4.5)
|
|
4(6.4)
|
1 (2.0)
|
|
|
1a
|
Dose modification
|
|
2 (40)
|
|
4 (18.2)
|
11 (17.7)
|
7 (11.1)
|
18 (6.9)
|
|
|
1b
|
Dose delaya
|
|
3 (60)
|
2 (66.7)
|
3 (13.6)
|
20 (32.3)
|
17 (27)
|
77(29.6)
|
|
|
2
|
Second line
|
Docetaxel
|
1 (20)
|
1 (33.3)
|
|
7 (11.3)
|
2 (3.2)
|
41 (15.8)
|
|
|
EOX
|
1 (20)
|
|
|
|
|
2 (0.8)
|
|
|
Capecitabine
|
|
|
3 (13.6)
|
|
|
5 (1.9)
|
|
|
CAPOX
|
|
|
|
|
|
1 (0.4)
|
|
|
FOLFIRI
|
|
|
|
|
|
1 (0.4)
|
|
|
2a
|
Dose modification[b]
|
|
|
|
|
2 (3.2)
|
|
10 (3.8)
|
|
|
2b
|
Dose delay
|
|
|
|
|
2 (3.2)
|
|
2 (0.8)
|
|
|
3
|
Third line
|
Capecitabine
|
1 (20)
|
|
|
|
|
1 (0.4)
|
|
|
FOLFIRI
|
|
|
|
|
|
1 (0.4)
|
|
|
Irinotecan
|
|
|
|
|
|
2 (0.8)
|
|
Abbreviations: EOX, epirubicin, oxaliplatin, and capecitabine; CAPOX, capecitabine
and oxaliplatin; CAP-CIS, capecitabine and cisplatin; FLOT, 5FU, leucovorin, oxaliplatin,
and docetaxel; FOLFIRI, 5FU, oxaliplatin, leucovorin, irinotecan.
b Dose modification in GCY(n = 6) [reason: poor performance status (n = 3), grade 3 diarrhea (n = 2), grade 4 thrombocytopenia (n = 1)].
aDose delay in GCY (n = 8) [reason: patients defaulted (n = 3), grade 3 neutropenia (n = 1), grade 3 diarrhea (n = 3), grade 4 thrombocytopenia (n = 1)].
Overall, 68% of patients could complete more than three cycles of chemotherapy. Various
treatment modalities in both groups have been compared and represented in [Table 3]. The percentage of patients receiving curative-intent chemotherapy was similar in
GCY and GCO. In patients receiving curative intent therapy, the choice of chemotherapy
regimen, dose modification, and dose delay were identical in both groups. The commonest
chemotherapy regimen used in the palliative setting was the EOX regimen in 59% of
GCY and 70% of GCO. A similar number of patients received first- and second-line palliative
chemotherapy in both groups.
The median duration of follow-up for all patients was 27 (range, 24–29) months. The
median OS of the entire population was 11 (range, 10–12) months. In GCY, the median
OS was not reached for patients treated with curative intent, and it was 13 months
for those treated with palliative intent. Similarly, in GCO on curative intent therapy,
the median OS was 22 months (range, 17.01–26.99), and of those treated with palliative
intent, the median OS was 9 months (7.87–10.14). Survival curves are shown in [Fig. 2].
Fig. 2 Kaplan–Meier survival estimate for overall survival of gastric cancer in young (GCY)
and gastric cancer in old (GCO) treated with curative and palliative intent.
Univariate analysis and multivariate analysis for OS in GCY is shown in [Table 4]. There was no statistically significant difference in the 2-year survival of the
variables analyzed. However, numerically, males had better 2-year OS than females
(47.4% vs. 22.1%), and GCY with intestinal histology had better 2-year OS than those
with diffuse histology (62% vs. 19%).
Table 4
Univariate and multivariate analyses for overall survival in GCY
|
S. No.
|
Parameter
|
N
|
Median OS (months)
|
Univariate analysis
|
Multivariate analysis
|
|
HR (95% CI)
|
p-Value
|
HR (95% CI)
|
p-Value
|
|
1
|
Gender (n = 33)
|
|
|
|
1.00
|
|
0.12
|
|
Male
|
11
|
13 (0.0-29.9)
|
Reference
|
|
Reference
|
|
|
Female
|
22
|
19(4.4-33.6)
|
1.00 (0.30-3.33)
|
|
0.22 (0.03-1.45)
|
|
|
2
|
ECOG (n = 33)
|
|
|
|
0.39
|
|
0.55
|
|
1
|
30
|
19.8(11.1-28.4)
|
Reference
|
|
Reference
|
|
|
>1
|
3
|
11.0 (6.2-15.8)
|
1.97 (0.42-9.35)
|
|
0.50 (0.05-4.82)
|
|
|
3
|
Histology (n = 30)
|
|
|
|
0.34
|
|
0.25
|
|
Intestinal
|
8
|
Not reached
|
Reference
|
|
Reference
|
|
|
Diffuse
|
22
|
19 (7.4-30.6)
|
1.88 (0.51-6.90)
|
|
3.70 (0.40-34.20)
|
|
|
4
|
Site of tumor (n = 33)
|
|
|
|
0.49
|
|
0.64
|
|
GEJ/cardia
|
9
|
Not reached
|
Reference
|
|
Reference
|
|
|
Fundus and body
|
6
|
11 (6.7-15.3)
|
3.07 (0.48-19.69)
|
|
3.89 (0.24-64.13)
|
|
|
Antrum and pylorus
|
18
|
19 (11.5-26.5)
|
1.71 (0.36-8.21)
|
|
1.99 (0.25-15.69)
|
|
|
5
|
GOO (n = 33)
|
|
|
|
0.16
|
|
0.07
|
|
Yes
|
7
|
Not reached
|
0.23 (0.03-1.79)
|
|
0.10 (0.01-1.22)
|
|
|
No
|
26
|
13 (5.6-20.4)
|
Reference
|
|
Reference
|
|
|
6
|
Stage (n = 33)
|
|
|
|
0.09
|
|
0.36
|
|
Nonmetastasis
|
13
|
35 (2.6-67.4)
|
Reference
|
|
Reference
|
|
|
Metastasis
|
20
|
19 (10.1-27.9)
|
3.09 (0.81-11.76)
|
|
2.53 (0.35-18.22)
|
|
|
7
|
Metastasis (n = 20)
|
|
|
|
0.23
|
|
0.43
|
|
Single
|
11
|
11 (2.4-19.6)
|
Reference
|
|
Reference
|
|
|
Multiple
|
9
|
Not reached
|
0.36 (0.07-1.81)
|
|
1.96 (0.36-10.57)
|
|
Abbreviations: ECOG, Eastern cooperative group; GEJ, gastroesophageal junction; GOO,
gastric outlet obstruction; OS, overall survival.
Discussion
Over the past few decades, we have seen significant changes in GC's biology, incidence,
and outcomes worldwide.[1]
[8]
[9] Change in lifestyle, addiction, and food habits, rampant use of antacids and proton-pump
inhibitors, early detection and treatment of Helicobacter pylori-induced gastritis, availability and increased accessibility of treatment and newer
therapies are a few reasons behind this change.[8]
[10]
[11] These reasons vary with a person's age, and the risk of developing cancer changes
with underlying genetic vulnerability and cumulative pressure from exposure to risk
factors throughout one's lifetime. Sparse data are available in GC patients less than
30 years of age. Our study reports the clinical characteristics, treatment, and survival
of GC patients less than 30 years of age compared to > 30 years of age.
Most studies used an age cut-off of less than 40 years for GCY. Our research's median
age of GC is one decade less than that seen in western countries (55 years vs. 68
years), so we defined GCY as less than 30 years.[2] Contrary to the GCO, the incidence of GCY is rising. Various studies reported the
incidence of GCY between 2 and 8%.[3]
[4]
[5]
[6] Our study had 4.6% of GCY among the registered GC cases, similar to that reported
in the literature.
In our study, GCY is more common in females (67%) than males. A higher female proportion
is the most common finding reported in the literature for GCY, indicating that sex
hormones, especially estrogen, may play an essential role in GCY development.[5]
[6] Zhou et al and Matsuyama et al showed that ER-beta expression rather than ERα expression
correlated with young age and advanced cancer stages in GCY.[12]
[13] For males, exposure to environmental risk factors, such as smoking and alcohol intake,
involves a sequence of preneoplastic lesions, contributing to increased GC incidence
later in life.
The proximal GC incidence increases in the developed world concordance with esophageal
cancers, suggesting that these might have similar risk factors and pathologies. However,
in India, the distal GC is still the most common, as reported in the literature.[14] In our study, antrum and pylorus were the most common sites in GCY (55%) and GCO
(60%).
Diffuse type gastric cancer (DGC) histology was more common in GCY than GCO (66.7%
vs. 41.7%, 0.001). This is in concordance with the various studies from the literature,
where DGC was more commonly detected in younger individuals.[6]
[7]
[15]
[16]
[17]
[18] This disproportion may be primarily genetically determined, mainly alterations in
the CDH1 gene, predisposing individuals to DGC at a younger age. Pathogenicity of DGC involves
multiple factors of cell signaling pathways, cell–cell adhesion, and H. pylori infection. The E-cadherin and cell-signaling pathways are vital in maintaining cell
integrity and normal cell function. The alterations in E-cadherin have been known
as a factor strongly associated factor with DGC.[6]
[19] None of the GCY patients had a family history of cancer, and due to logistics, genetic,
and molecular studies were not done in our patients.
The difference in the stage at presentation between GCY and GCO was found in most
studies. GCY mainly presented with locally advanced and node or distant metastatic
disease.[5]
[6]
[17]
[18]
[20] In addition to the aggressive diffuse histology type, delay in diagnosis is also
a reason for the advanced stage at presentation. The main reason for the delay in
diagnosis is that GC was not considered a differential diagnosis in young patients
presenting with gastrointestinal symptoms and was not assigned to routine endoscopic
screening.[6] Despite no delay in the diagnosis in our study, 60% of GCY had metastases at presentation,
indicating the aggressive biology of the disease. Though GCY presented in the advanced
stage, there was no difference in the literature's incidence of multiple site metastases
in GCY and GCO. However, in our study, the incidence of multiple metastases (two or
more sites) was more common in GCY than in the GCO.
While some studies demonstrated poorer outcomes in young patients, the majority reported
a better prognosis than older individuals, and some still have no differences in survival
based on age.[6]
[21]
[22]
[23]
[24] Even though GCY had more diffuse-type histology and aggressive presentation, they
had better performance status, less comorbidity, and similar[18] tolerance to chemotherapy, which resulted in similar survival compared to GCO. Our
study also showed no difference in survival between the two groups. In our research
(GCY), male sex and intestinal type histology had better survival but did not reach
statistical significance due to the small sample size and short follow-up.
The strength of this study is that we are comparing the data with GCO from the same
population. Limitations are retrospective data, small sample size, and different comparison
periods. Unique challenges in GCY, such as fertility preservation and psychosocial
problems, could not be analyzed as we did not have the data.
Conclusion
GCY is more common in females and has aggressive diffuse-type histology with multiple
metastases than GCO. Even though GCY had more diffuse-type histology and aggressive
presentation, they had better performance status, fewer comorbidities, and similar
OS compared to GCO. A separate registry for this unique subset of patients to study
the detailed genetic factors, etiology, clinical characteristics, treatment adherence,
sexual health, and psychosocial problems would help understand the pathogenesis, treatment
response, and outcomes
