Gastric cancer - RJME

Romanian Journal of Morphology and Embryology 2009, 50(2):185–194

ORIGINAL PAPER Gastric cancer: correlation between clinicopathological factors and survival of patients (II) DANIELA LAZĂR1), SORINA TĂBAN2), I. SPOREA1), ALIS DEMA2), MĂRIOARA CORNIANU2), ELENA LAZĂR2), A. GOLDIŞ1), C. VERNIC3) 1)

Department of Gastroenterology and Hepatology 2)

3)

Department of Pathology

Department of Medical Informatics and Biostatistics

“Victor Babeş” University of Medicine and Pharmacy, Timisoara

Abstract

Purpose: The purpose of this study was to identify the clinicopathological factors that influence the prognosis of patients having undergone surgery for gastric cancer. The analysis of the potential prognosis factors has included in this study parameters concerning the patient (sex, age), as well as parameters related with the tumor (macroscopic aspect of the tumor according to the Borrman’s classification, histological type according to the WHO and Lauren’s classifications, degree of tumor differentiation, tumor location, stage of disease, pT and pM parameters according to the TNM classification of AJCC/UICC). Material and Methods: From the total number of 265 patients (186 men and 79 women) diagnosed with gastric cancer in the period 1998–2002, 61 operated patients were selected. On this group, we performed a prospective study regarding the evolution and aggressiveness of gastric cancer, on a duration of 5 years. Survival time was calculated from the month of the surgical intervention until the month of death or confirmation of survival, and survival rate was represented by the percentage of survivals at the end of the observed interval (in years and months). Results: The studied group consisted of 61 patients (43 men and 18 women), with ages between 30 and 80 (average age = 59.34 years). We have identified five papillary adenocarcinomas (8.2%), 28 tubular adenocarcinomas (46%), 17 “signet-ring” cell carcinomas (27.8%), eight mucinous adenocarcinomas (13.1%), and three undifferentiated or anaplastic carcinomas. Most gastric carcinomas examined were included in the category of poorly differentiated carcinomas (63.9%). According to the Lauren’s classification, we have identified 38 intestinal type gastric carcinomas (62.3%), 17 diffuse type carcinomas (27.9%), and six mixed carcinomas (9.8%). We have identified aspects of lymphovascular invasion in 38 cases (62.3%). Conclusions: “Signet-ring” cells carcinomas, and the anaplastic ones, prove in our study to be extremely aggressive histological forms, characterized through low rates of survival. We remarked a significant correlation between the degree of tumor differentiation and survival of patients, the values recorded being significantly lower in medium and poorly differentiated carcinomas (p = 0.00871194 FS). Average survival, calculated in months, is significantly lower in patients with diffuse type carcinomas (11.3 months), in comparison with patients presenting intestinal type carcinomas (20.4 months) (p = 0.0415 S). There is a direct proportional relationship between the lymphovascular invasion and the number of positive lymph nodes. Survival after 5 years decreases significantly in the presence of lymphovascular invasion. Keywords: gastric cancer, clinicopathological factors, survival.

Introduction Gastric cancer represents one of the most frequent neoplasias. Although its incidence decreased over the last few decades in industrialized countries, it still represents nowadays a major cause of death through cancer throughout the world. Prognosis of gastric cancer is generally reserved. The low survival rate is due to the delay in diagnosis, most cases being diagnosed in an advanced stage, as well as to the frequent local recurrences. The purpose of this study was to identify clinicopathological parameters that influence the prognosis of patients operated for gastric cancer. Literature studies offer controversial data related to factors that influence prognosis in gastric cancer. The analysis of potential prognosis factors has included in this study parameters related to the patient

(sex, age) as well as parameters related to the tumor (macroscopic aspect of tumor, according to the Borrmann’s classification, histological type according to the WHO and Lauren’s classifications, degree of tumor differentiation, tumor location, stage of disease, pT, pN, and PM parameters according to the TNM classification of AJCC/UICC). Material and Methods From the total number of 265 patients (186 men and 79 women) clinically and histopathologically diagnosed with gastric cancer in the period between 1998 and 2002, 67 patients were selected, who underwent surgery for this pathological condition in the Departments of Surgery of the Emergency County Hospital of Timişoara. On this group, we performed a prospective study regarding the evolution and aggressiveness of

Daniela Lazăr et al.

186

gastric cancer, over a period of 5 years. Surgical interventions performed, with curative or palliative intentions, were not preceded by chemotherapy or radiotherapy. The patients or their families were contacted periodically, on the phone, or through medical letters, at intervals of 6 months, the survival being monitored over a variable period between 1 and 68 months. Patients who died in the period after the surgery, through various complications, or due to other conditions, were excluded from the study. Clinical and morphological (macroscopic and microscopic) data were gathered for each case. Gastric carcinomas were classified and interpreted according to the evaluation protocol recommended by the American Joint Committee on Cancer (AJCC) and International Union against Cancer (IUCC). Survival time was calculated from the month of surgery until the time of death or confirmation of survival, and survival rate was represented by the percentage of survivals at the end of the observed interval (in years and months). From the total of cases included in the prospective study, six patients died at intervals variable between 7 and 26 months, due to other medical causes, being excluded from the study. Statistical analysis was performed using the EpiInfo 6.04, Epi 3.2.2 and OpenEpi and consisted in computing the frequency counts and percentages for the qualitative variables, the means and standard deviations for the quantitative variables. The comparison of the percentages and the means was performed using the chi square test and the unpaired Student t-test. For statistical analysis, p values of less than 0.05 were considered significant, and p values of less than 0.01 were considered very significant. Results The final group consisted of 61 patients (43 men and 18 women) with ages between 30 and 80 (average age = 59.34 years). The main clinicopathological features of cases of gastric cancer investigated are presented in Table 1. For the histopathological evaluation of cases, we used the WHO histological classification of gastric carcinomas, which includes the following categories: ▪ Papillary adenocarcinoma;

▪ Tubular adenocarcinoma; ▪ Mucinous adenocarcinoma; ▪ “Signet-ring” cell carcinoma; ▪ Adenosquamous carcinoma; ▪ Small cell carcinoma; ▪ Non-differentiated carcinoma ▪ Other forms of carcinoma. Malignant “signet-ring” cells that infiltrate the gastric wall induce a dense stromal fibrous reaction. Table 1 – Clinicopathological features of gastric cancers studied Clinicopathological factors

No. of cases

Men Women Average age (min.–max.) [years] Antrum Body Location Pangastric Eso-cardia Gastric stump Early carcinoma Advanced carcinoma I II Borrmann III IV pTis/T1/T2/T3/T4 pN0/N1/N2/N3 pM0/M1

43 18 59.34 (30–80) 31 15 10 2 3 5 56 5 20 22 9 4/6/7/21/23 18/16/23/4 47/14

In our case studies, we identified five papillary adenocarcinomas (8.2%) (Figure 1), 28 tubular adenocarcinomas (46%) (Figure 2), 17 “signet-ring” cell carcinomas (27.8%) (Figures 3 and 4), and eight mucinous adenocarcinomas (13.1%) (Figure 5). In three cases (4.9%) malignant neoplastic proliferations with solid pattern (Figure 6), with expressed cellular and nuclear pleomorphism and with elevated mitotic activity, were classified as undifferentiated or anaplastic carcinomas. In order to diagnose and classify these tumors correctly we performed histochemical reactions with anti-panCK antibodies, the MNF116 clone (Figure 7), and anti-CK7, the OV–TL 12/30 clone. Distribution of cases according to age and sex of patients is depicted in Table 2.

Table 2 – Distribution of histological forms according to patients’ age and sex Age groups [years] TA 1 31–40 – 41–50 5 51–60 8 61–70 3 71–80 17 Total no. of patients (39.5%) (%)

PA – 2 – 2 – 4 (9.3%)

Men SRCC 2 3 4 3 – 12 (27.9%)

MA 1 1 2 2 1 7 (16.3%)

AC – – 2 – 1 3 (7%)

TA 1 – 2 3 5 11 (61.1%)

PA – – 1 – – 1 (5.5%)

Women SRCC – 1 1 3 – 5 (27.9%)

MA – – – 1 – 1 (5.5%)

AC – – – – – –

TA – tubular adenocarcinoma; PA – papillary adenocarcinoma; SRCC – “signet-ring” cell carcinoma; MA – mucinous adenocarcinoma; AC – anaplastic carcinoma.

Tubular and papillary adenocarcinomas were noted in patients from all age groups, being more frequently noted for the age groups of 51–60 years and 61–70 years, in both sexes. We remarked the development of

the “signet-ring” cell carcinoma in young patients and in medium age patients (two cases for the 31–40 years group, three cases for the 41–50 years group, and four cases for the 51–60 years group). Mucinous adenocarci-

Gastric cancer: correlation between clinicopathological factors and survival of patients (II)

noma was noted predominantly in males (16.3%). Also, undifferentiated carcinoma was observed only in men. In women, the most frequent histological type was tubular adenocarcinoma (61.1%), followed by the “signet-ring” cell carcinoma (27.9%). We noted five cases of carcinomas with “signet-ring” cells in female patients from the age groups 41–50 years, 51–60 years, and 61–70 years. According to the histological type, we noted the following average values of survival: 19.6 months for

187

patients with papillary adenocarcinomas; 23.7 months for patients with tubular adenocarcinomas; 21 months for patients with mucinous adenocarcinomas; 7.6 months for patients with undifferentiated carcinomas; 10.7 months for patients with “signet-ring” cell carcinomas. We remarked close values of survival for patients with various subtypes of adenocarcinomas; however, carcinomas with “signet-ring” cells and anaplastic carcinomas prove to be extremely aggressive forms, characterized by reduced rates of survival (Figure 8).

Figure 1 – Papillary adenocarcinoma (HE stain, ×200).

Figure 2 – Tubular adenocarcinoma (HE stain, ×200).

Figure 3 – “Signet-ring” cell carcinoma (HE stain, ×200).

Figure 4 – Alcianophil secretion in the cytoplasm of “signet-ring” cells (AA–PAS stain, ×400).

Figure 5 – Colloid carcinoma (HE stain, ×200).

Figure 6 – Anaplastic carcinoma (HE stain, ×400).


188

19,6

PA

23,7

TA 21

MA 10,7

SRCC 7,6

AC 0

5

10

15

20

25

Average survival (months)

Figure 7 – Anaplastic pan CK (DAB, ×400).

carcinoma.

Immunoreaction

We noted a relatively low rate of survival in patients with papillary, tubular, and mucinous adenocarcinoma. At the end of the interval of 5 years, we noted survival rates of 20% (one case) for papillary adenocarcinoma, 25% (seven cases) for tubular adenocarcinoma, and 23% for mucinous adenocarcinoma (two cases). Two patients with anaplastic carcinoma died in the first year of evolution, and the third in the next year. From the 17 patients with “signet-ring” cell carcinomas, 12 died in the first year, survival rate at 5 years being 0% (Figure 9). According to the extension of glandular differentiation in the tumor mass and to cell anomalies, gastric adenocarcinoma is classified in three types: well-differentiated (G1), moderately differentiated (G2), and poorly differentiated (G3). This classification is applied to papillary, tubular, and mucinous adenocarcinomas. The carcinoma with “signet-ring” cells is by its nature a poorly-differentiated form (Figure 10). In the studied cases we identified only two welldifferentiated carcinomas (3.3%) and 20 moderatelydifferentiated carcinomas (32.8%) (Figure 11). Most gastric carcinomas examined presented a weak formation of glands and significant cellular anomalies,

Figure 8 – Survival of patients according to histological type of gastric cancer.

being classified in the category of poorly-differentiated carcinomas (63.9%) (Figure 12). Distribution of tumoral grading according to patients’ age and gender is depicted in Table 3. Table 3 – Degree of differentiation according to patients’ age and gender Age groups [years] 31–40 41–50 51–60 61–70 71–80 Total [%]

Men G1

G2

Women G3

G1

G2

G3

1 – 3 1 – – – 1 5 – – 1 – 6 7 – 2 2 – 5 10 – 2 5 – 1 4 – 3 2 1 13 29 1 7 10 (2.3%) (30.2%) (67.5%) (5.5%) (38.9%) (55.6%)

We noted the development of well-differentiated carcinomas in young patients, in the age group of 31–40 years. Medium-differentiated carcinomas were noted in patients with medium and old ages. A great number of poorly differentiated carcinomas (G3) were noted in young subjects, from the age group of 31–40 years, and with medium ages (41–50 years).

120 100 80 60

2

40

20

20 0 0

PA

1 year

TA

Histological type

2 yrs

3 yrs

MA

4 yrs

5 yrs

SRCC

AC

1 year 2 years 3 years 4 years 5 years

PA (papillary adenocarcinoma) 60% TA (tubular adenocarcinoma) 46.4% 50% MA (mucinous adenocarcinoma) SRCC (“signet-ring” cell carcinoma) 29.4% AC (anaplastic carcinoma) 33.3%

60%

20%

20%

35.7%

32.1%

25%

25%

25%

25%

25%

25%

11.7%

5.9%

0%

0%

0%

0%

0%

0%

39

20%

Figure 9 – Survival at 5 years according to the histological type [%].

G1 Carcinoma

G2 Carcinoma

G3 Carcinoma

Figure 10 – Distribution of gastric carcinomas according to degree of tumor differentiation.


Figure 11 – Moderately differentiated adenocarcinoma (G2), with cribriform glandular structures (HE stain, ×400).

Survival at 5 years for patients with G1 carcinomas was of 100%, both being alive at the end of the period recorded. From the patients with G2 carcinomas, 15 have died (11 patients in the first year, a second patient in the second year, one patient in the third year, and two patients in the fourth year), and the average rate of survival was of 25%. In the group of G3 carcinomas we noted a much reduced survival over 5 years (7.7%), 36 patients dying during this period (27 patients in the first year, eight patients in the second year, and one patient in the third year). The greatest number of deaths was noted in the first year of post-surgery evolution, both for medium-differentiated and for poorlydifferentiated carcinomas (Figure 13). Calculating the average survival rate in months, we obtained the following values: for patients with G1 carcinomas – 60 months; for patients with G2 carcinomas – 22.3 months; for patients with G3 carcinomas – 11.2 months. We already noted the significant correlation between the degree of tumor differentiation and survival of patients with gastric cancers, the values recorded being significantly lower in mediumand poorly-differentiated carcinomas (p = 0.00871194) (Figure 14).

Figure 12 – Poorly differentiated (G3) (HE stain, ×400).

189

adenocarcinoma

According to the structural aspects, the Lauren’s classification of gastric carcinomas includes two major categories of tumors: ▪ Intestinal-type carcinomas, histologically characterized by big glandular structures, well-defined, lined with relatively well-differentiated malignant cells, sometimes with papillary tendency, resembling colon adenocarcinomas. Malignant cells are column-like or cuboidal, presenting a basally situated nucleus in the well-differentiated types. In this type of carcinoma, cells do not contain, or contain reduced quantities of mucines; however, in the lumen of glands mucinous secretion can be identified. Pseudo-glandular structures are of various dimensions, sometimes cystically dilated. ▪ Diffuse-type carcinomas, represented by isolated malignant cells, discohesive or arranged in small groups, of the “signet-ring” cells type, or undifferentiated, with a marked infiltrative characterristic. Tumor proliferation does not induce significant architectural modifications of the remaining gastric glands. There is usually a well-represented, fibrous, dense stroma, with a discrete inflammatory infiltrate.

120 100 80 60 40

11,2

G3

20

22,3

G2

0 0

1 year

2 yrs

G1

4 yrs

G2

Degree of differentiation 1 year

G1 G2 G3

3 yrs

60

5 yrs

G3

2 years

3 years

4 years

5 years

100% 45%

100% 40%

100% 35%

100% 25%

100% 25%

30.7%

10.2%

7.7%

7.7%

7.7%

Figure 13 – Survival at 5 years according to the degree of tumor differentiation.

G1 0

20

40

60


Figure 14 – Survival of patients according to the degree of tumor differentiation.

190


Rarely, tumors present a mixed characteristic, combining the aspects of intestinal carcinoma with those of the diffuse carcinoma (Figure 15). According to the Lauren’s classification, we identified 38 intestinal-type gastric carcinomas (62.3%), 17 diffuse-type carcinomas (27.9%), and six mixed carcinomas (9.8%) (Figure 16). In our study, intestinal-type and diffuse type carcinomas develop more frequently in males (68.4% and 82.3%) (Table 4). Table 4 – Lauren’s classification and clinicopathological factors in gastric cancers Clinicopathological Intestinal Diffuse Mixed factors type (n = 38) type (n = 17) type (n = 6) 26 14 3 Men 12 3 3 Women Average age 62.84 ± 10.18 53.29 ± 11.23 54.33 ± 12.69 [years] 20 8 3 Antrum 10 3 2 Body 4 5 1 Pangastric 2 – – Eso-cardia 2 1 – Gastric stump 5 – – Early carcinoma Advanced 33 17 6 carcinoma 2 2 1 I 14 4 2 II Borrmann 15 4 3 III 2 – – IV 28 – – TA 5 – – PA 4 – 4 MA – 15 2 SRCC 1 2 – AC 2 – – G1 20 – – G2 16 17 6 G3 Average survival 20.4 11.3 15.8 [months] 21.1% 5.9% 16.6% Survival at 5 years

The average age in diffuse type carcinoma (53.29 ± 11.23 years) is significantly lower than the average age of patients with intestinal-type carcinomas

(62.84 ± 10.18 years) (p = 0.00288). Antral location is predominant in intestinal-type carcinomas (52.6%). For diffuse-type carcinomas, we noted mostly extended carcinomas, involving the entire stomach (29.4%). We noted an important correlation between the level of tumoral invasion and the Lauren’s type of carcinoma. Thus, all early gastric carcinomas are represented histologically by intestinal-type carcinomas (100%). In intestinal-type carcinomas, the II and III Borrmann’s forms are predominant (36.8% and 39.5%). The most numerous diffuse tumors are represented by infiltrative tumors, of the IV Borrmann’s type (41.2%). Histologically, 73.7% of intestinal carcinomas are tubular adenocarcinomas and much more rarely papillary adenocarcinomas (13.1%), mucinous (10.5%), or anaplastic carcinoma (2.7%). Carcinoma with “signet-ring” cells displays in a proportion of 88.2% the aspect of the diffuse type carcinoma. In this category, we did not include any cases of adenocarcinomas. The intestinal-type consists of well-differentiated carcinomas (5.3%), medium-differentiated (52.6%), and poorly-differentiated (42.1%). In a proportion of 100% of cases, diffuse-type carcinomas consist of carcinomas with “signet-ring” cells. Average survival rate, calculated in months, is significantly lower in patients with diffuse-type carcinomas (11.3 months) in comparison with patients presenting intestinal-type carcinomas (20.4 months) (p = 0.0415). In mixed-carcinomas, we noted a medium value of survival rate (15.8 months) (Figure 17). In intestinal-type carcinomas, we noted 30 deaths, the survival rate over 5 years being 21.1% (Figure 18). Due to the great number of patients deceased (16 patients from a total of 17), survival rate over 5 years in diffuse-carcinoma was of 5.9%, significantly lower (p = 0.0312). In mixed-type carcinomas, we noted an average value of survival of 16.6%. For all forms of carcinomas, the greatest number of deaths was recorded in the first year after surgery. Transversal sections through the gastric wall, including the tumor and perigastric tissues, were examined morphologically in order to identify the vascular, lymphatic, and blood invasion. 6

17

38 Intestinal type carcinoma Diffuse type carcinoma Mixed type carcinoma

Figure 15 – Gastric carcinoma of a non-classifiable type (mixed) (HE stain, ×200).

Figure 16 – Lauren’s carcinomas.

classification

of

gastric


191

120 100 80

15,8

Mixed Type Cc

40

11,3

Diffuse Type Cc

60

20,4

20 0

Intestinal Type Cc

0

0

5

10

15

20

25


and

In assessing this morphological parameter, we made no distinction between lymphatic and blood vessels, noting the presence/absence of lymphovascular invasion for each case. On the HE staining, we interpreted a vascular invasion only in the presence of the welldefined layer of endothelial cells, situated around malignant cells (Figure 19). From the group of 61 gastric carcinomas studied, we identified aspects of lymphovascular invasion in 38 cases, representing 62.3%. Lymphovascular invasion related to various clinicopathological factors is presented in Table 5. Men patients develop much more frequently tumors with lymphovascular invasion (73.7% of cases). We did not note the existence of a relationship between the average age of patients and the vascular invasion, the two values obtained being very close (p = 0.631487). Also, we did not obtain a correlation between the location of tumors and this parameter, with the exception of cardial location. Both eso-cardial tumors were accompanied by intravascular carcinomatous emboli. In our study, early gastric carcinomas did not present aspects of lymphovascular invasion. Tumors with vascular involvement consisted in a proportion of 100% of advanced carcinomas. Among macroscopically types of gastric cancer, only the IV Borrmann’s type (infiltrative) was more frequently observed in tumors with lymphovascular invasion (18.4%), in comparison with tumors without lymphovascular invasion (8.7%). The other macroscopically forms were observed in similar proportions in the other two categories of tumors. We observed a significant correlation between the Lauren classification of gastric carcinomas and the presence of lymphovascular invasion. Thus, 42.1% of tumors with intravascular carcinomatous emboli are represented by diffuse type carcinomas. Very rarely, in only 4.3% of cases, diffuse type carcinomas do not produce lymphovascular invasion. In nearly all anaplastic carcinomas and carcinomas with “signet-ring” cells, we identified lymph vessels with carcinomatous emboli, demonstrating the strongly metastasizing characteristic of these histopathological forms of gastric cancer. Most tumors without lympho-

3 yrs

Diffuse Type Cc

4 yrs

5 yrs

Mixed Type Cc

Lauren’s classification

1 year

2 years

3 years

4 years

5 years

Intestinal cc.

39.5% 35.3% 33.3%

28.9% 11.7% 16.6%

26.3% 5.9% 16.6%

21.1% 5.9% 16.6%

21.1% 5.9% 16.6%

Diffuse cc.

Lauren’s

2 yrs

Intestinal Type Cc

Mixed cc.

Figure 17 – Survival of patients classification of gastric carcinomas.

1 year

Figure 18 – Survival at 5 years classification of gastric carcinomas.

and

Lauren’s

vascular invasion consist of tubular adenocarcinomas (74% of cases). Also, there is a close correlation between the degree of tumor differentiation and the vascular invasion. 84.2% of tumors with lymphohematogenous invasion are poorly-differentiated carcinomas. Table 5 – Relation between lymphovascular invasion and clinicopathological factors in gastric cancers Tumors with Clinicopathological lymphovascular factors invasion (n = 38) 28 Men 10 Women 58.79 ± 10.75 Average age [years] 18 Antrum 9 Body 7 Pangastric 2 Eso-cardia 2 Gastric stump – Early carcinoma 38 Advanced carcinoma 4 I 12 II Borrmann 15 III 7 IV 17 Intestinal carcinoma 16 Diffuse carcinoma 5 Mixed carcinoma 11 TA 3 PA 5 MA 16 SRCC 3 AC 0/6/32 G1/G2/G3 0/0/3/13/22 pTis/T1/T2/T3/T4 4/12/18/4 pN0/N1/N2/N3 25/13 pM0/M1 Average survival rate 11 [months] 7.9% Survival at 5 years

Tumors without lymphovascular invasion (n = 23) 15 8 60.26 ± 12.77 13 6 3 – 1 5 18 1 8 7 2 21 1 1 17 2 3 1 – 2/14/7 4/6/4/8/1 14/4/5/0 22/1 28.1 30.4%

In relation with the level of tumoral invasion, the vascular permeation was observed much more frequently in T4 carcinomas (57.9%) and T3 (34.2%).


192

There is a direct proportional relationship between tumoral lymphatic invasion and the number of positive lymph nodes. Tumors with lymphovascular invasion are represented by pN0 carcinomas – 10.5%; pN1 carcinomas – 31.6%; pN2 carcinomas – 47.4%; pN3 carcinomas – 10.5%. In tumors without aspects of vascular permeation, we identified 60.9% pN0 carcinomas, 17.4% pN1 carcinomas, 1.7% pN2 carcinomas, and 0% pN3 carcinomas. Lymph-blood invasion is correlated with the presence of distant metastases. Tumors with aspects of vascular invasion are associated in a rate of 34.2% with distant metastases, unlike carcinomas without vascular

permeation, with distant metastases in only 4.3% of cases. Survival at 5 years decreases significantly in the presence of lymphovascular invasion. Survival rate is of only 7.9% in carcinomas with intravascular tumoral emboli, in comparison to survival at 5 years of 30.4% in carcinomas without lymphovascular invasion (p = 0.041442) (Figure 20). Survival of patients, calculated in months, showed obviously different values. In tumors with lymphohematogenous invasion, patients survived for 11 months, on average, unlike patients with tumors without lymphovascular invasion, who survived, on average, 28.1 months (Figure 21). 120 100 80 60 40 20 0 0

1 year

2 yrs

3 yrs

Lymphovascular Invasion + Tumor invasion Lymphovascular invasion + Lymphovascular invasion -

Figure 19 – Carcinomatous embolus in a lymph vessel (HE stain, ×200).

28,1

Lymphovascular invasion 11

Lymphovascular invasion + 0

10

20

30


Figure 21 – Survival of patients according to the lymphovascular invasion.

Discussion In our case study, the predominant histological type was tubular adenocarcinoma, observed in nearly half of all patients. It is to be noted that almost 1/3 of all patients presented carcinomas with “signet-ring” cells. Among the other histological types observed were mucinous adenocarcinomas (13.1%), papillary adenocarcinomas (8.2%), and undifferentiated carcinomas (4.9%). We observed the development of the carcinoma with “signet-ring” cells in young patients, aspect described by numerous studies in literature. We noted nearly similar values of survival at 5 years for patients with various types of adenocarcinomas, 25% for tubular adenocarcinoma, 23% for mucinous adenocarcinoma, and 20% for papillary adenocarcinoma; however, carcinomas with “signet-ring” cells and anaplastic carcinomas prove to be extremely aggressive forms, as shown by other studies from

4 yrs

5 yrs

Lymphovascular invasion -

1 year

2 years

3 years

4 years

5 years

26.3% 56.5%

10.5% 43.5%

7.9% 39.1%

7.9% 30.4%

7.9% 30.4%

Figure 20 – Survival at 5 years according to the lymphovascular invasion.

literature [1, 2], characterized by reduced rates of survival, none of the patients with these histological forms surviving at 5 years. All patients presenting undifferentiated carcinomas died during the first two years after surgery. In our study, most gastric carcinomas examined were classified in the category of poorly differentiated carcinomas (63.9%). A great number of poorly differentiated carcinomas (G3) were noted in young patients, from the age group of 31–40 years, and with medium ages (41–50 years). Concerning histology related to age groups, numerous studies have shown that significantly more elders present well- or moderatelydifferentiated tumors, while significantly more young patients present poorly-differentiated carcinomas and carcinomas with “signet-ring” cells. Nakamura T et al. [3] analyzed the histological types in the case of early GC in elderly patients, and reported that 45.5% of carcinomas were well-differentiated, as opposed to the much greater incidence of poorly-differentiated neoplasms recorded in young patients. Concerning the degree of cellular differentiation, literature data proved the most favorable prognosis in the case of well-differentiated tumors, well-differentiated neoplasms having a significantly better survival rate as compared with those poorly differentiated [4–8]. These aspects were also noted in our group study, demonstrating a significant correlation between the degree of tumor differentiation and survival of patients with gastric cancers, the values recorded being significantly lower in medium- and poorly-differentiated carcinomas.


In our case study, the intestinal type of gastric cancer was predominant (62.3%). The average age in diffuse-type carcinoma is significantly lower than the average age of patients with intestinal-type carcinomas (43 vs. 62.8 years), which concurs with literature data that describe the greater frequency of the diffuse type of GC in younger ages [4, 9–12]. Antral location is predominant in intestinal-type carcinomas (52.6%). For diffuse-type carcinomas, we noted mostly extended carcinomas, involving the entire stomach (29.4%). We observed an important correlation between the level of tumoral invasion and the Lauren’s type of carcinoma [13]. Thus, all early gastric carcinomas are represented histologically by intestinal-type carcinomas (100%). In intestinal-type carcinomas, the II and II Borrmann’s forms are predominant (36.8% and 39.5%). The most numerous diffuse tumors are represented by infiltrative tumors, of the IV Borrmann’s type (41.2%). Histologically, 73.7% of intestinal carcinomas are tubular adenocarcinomas. The carcinoma with “signet-ring” cells displays in a proportion of 88.2% the aspects of the diffuse-type carcinoma. The intestinal type is composed of well-differentiated carcinomas (5.3%), medium-differentiated carcinomas (52.6%), and poorly-differentiated carcinomas (42.1%). In 100% of cases, diffuse-type carcinomas consist of carcinomas with “signet-ring” cells. Average survival, calculated in months, is significantly lower in patients with diffuse-type carcinomas, in comparison with patients presenting intestinal-type carcinomas (11.3 vs. 20.4 months). The incidence of lymphovascular invasion in gastric cancer was observed in a percentage of 5.4–86% of cases, the smallest incidence being reported in patients with tumors without lymphonodal metastases [14]. Dicken BJ et al. [15] described the presence of lymphovascular invasion in 59.6% of patients with gastric cancer in whom curative resection of tumor was performed. Numerous studies reported that lymphovascular invasion represents an independent risk factor for long-term survival and for the development of lymph node metastases, suggesting the fact that lymphovascular invasion constitutes a clinical marker of tumoral aggressiveness [16–18]. This observation was sustained by the study of Hyung WJ et al. [19], who demonstrated that lymphovascular invasion represents a non-favorable prognosis indicator, independent from the clinicopathological factors, in gastric cancer without lymphonodal metastases, the study concluding that the lymphovascular invasion could offer useful information for the prognosis and clinical management of this subset of patients. Recently, Kooby DA et al. [20] showed that lymphovascular invasion in N0 patients represents an independent predictor of non-favorable prognosis, indicating the presence of more aggressive tumors, independent from the tumoral dimension and the depth of invasion [5, 21]. The study of Dicken BJ et al. [15], including all patients with gastric cancer, no matter the lymphonodal status, demonstrated that the status N, M, and lymphovascular invasion is associated independently with the survival of patients with resected gastric neoplasm.

193

These results are supported by the data obtained by Talamonti MS et al. [22], which recently showed that, besides other clinicopathological factors, lymphovascular invasion is associated with a period of absence of recurrences. This study reported a survival rate at 5 years of 26.2% in patients with lymphovascular invasion, compared to 49.9% in the case of patients without invasion. In the study of Dicken BJ et al. (346), a significant correlation was demonstrated between the lymphovascular invasion, status T and status N respectively, the survival of patients presenting invasion being signifycantly lower in comparison with patients in whom no lymphovascular invasion was noted (13.9% vs. 55.9%). From the group of 61 gastric carcinomas studied we identified aspects of lymphovascular invasion in 38 cases, representing 62.3%. Male patients develop much more frequently tumors accompanied by lymphovascular invasion (73.7% of cases). We did not observe the existence of a relationship between the average age of patients and vascular invasion, the two values obtained being nearly similar. Also, we did not obtain a correlation between the location of tumors and this parameter, except for the cardial location, in which the presence of lymphovascular invasion was detected. Tumors with vascular involvement consisted in a percentage of 100% of advanced carcinomas. The IV Borrmann’s type (infiltrative) was more frequently observed in tumors with lymphovascular invasion (18.4%), in comparison to tumors without lymphovascular invasion (8.7%). Most diffuse type carcinomas produce lymphovascular invasion (95.7%). In nearly all anaplastic carcinomas and carcinomas with “signetring” cells, we identified lymph vessels with carcinomatous emboli; 84.2% of tumors with lymphohematogenous invasion are poorly differentiated carcinomas. Vascular invasion was observed much more frequently in T4 carcinomas (57.9%) and T3 (34.2%), correlating as such with the depth of tumoral invasion. 89.5% from the tumors with lymphovascular invasion are associated with lymphonodal metastases. There is a direct proportional relationship between the lymphovascular invasion and the number of positive lymph nodes. The lymphohematogenous invasion is also correlated with the presence of distant metastases (34.2% vs. 4.3%). Survival at 5 years diminishes significantly in the presence of lymphovascular invasion. The rate of survival is of only 7.9% in carcinomas with intravascular tumoral emboli, in comparison with survival at 5 years of 30.4% in carcinomas without lymphovascular invasion. Conclusions According to the results of the study, carcinomas with “signet-ring” cells, and anaplastic carcinomas, prove to be extremely aggressive histological forms, characterized by reduced rates of survival, none of the patients with these histological forms surviving at 5 years. We noted a significant correlation between the degree of tumor differentiation and survival of patients, the values recorded being significantly lower in medium and poorly differentiated carcinomas (p = 0.00871194).

194


In the studied group, the intestinal type of gastric cancer was predominant (62.3%). The data obtained revealed that the average age in diffuse type carcinoma ((53.29 ± 11.23 years) is significantly lower than the average age of patients presenting intestinal type carcinomas (62.84 ± 10.18 years) (p = 0.00288). Antral location is predominant in intestinal type carcinomas (52.6%), while diffuse type tumors are represented most frequently by extended carcinomas, involving the entire stomach (29.4%). In 100% of cases, diffuse type carcinomas consist of carcinomas with “signet-ring” cells. The average survival, calculated in months, is significantly lower in patients with diffuse-type carcinomas (11.3 months) compared to patients presenting intestinal type carcinomas (20.4 months) (p = 0.0415). The aspects of lymphovascular invasion were present in 62.3% of the cases studied. According to our data, there is a directly proportional relationship between the lymphovascular invasion and the number of positive lymph nodes. The lymphohematogenous invasion is correlated with the diffuse, infiltrative type of gastric cancer, with the depth of invasion, the number of positive lymph nodes, and the presence of distance metastases. Survival at 5 years diminishes significantly in the presence of lymphovascular invasion. References [1] PARK J. M., JANG Y. J., KIM J. H., PARK S. S., PARK S. H., KIM S. J., MOK Y. J., KIM C. S., Gastric cancer histology: clinicopathologic characteristics and prognostic value, J Surg Oncol, 2008, 98(7):520–525. [2] LI C., KIM S., LAI J. F., HYUNG W. J., CHOI W. H., CHOI S. H., NOH S. H., Advanced gastric carcinoma with signet ring cell histology, Oncology, 2007, 72(1–2):64–68. [3] NAKAMURA T., YAO T., NIHO Y., TSUNEYOSHI M., A clinicopathological study in young patients with gastric carcinoma, J Surg Oncol, 1999, 71(4):214–219. [4] GORBUNOV E. A., WECHSLER J., STAŠEK T., TOMIN A. S., SOLODUN L. D., Significant prognostic factors in 283 patients after surgery for adenocarcinoma of the stomach, Scripta Medica (Brno), 2005, 78(1):3–16. [5] DICKEN B. J., SAUNDERS L. D., JHANGRI G. S., DE GARA C., CASS C., ANDREWS S., HAMILTON S. M., Gastric cancer: establishing predictors of biologic behavior with use of population-based data, Ann Surg Oncol, 2004, 11(6):629–635. [6] KIM D. Y., JOO J. K., RYU S. Y., PARK Y. K., KIM Y. J., KIM S. K., Clinicopathologic characteristics of gastric carcinoma in elderly patients: a comparison with young patients, World J Gastroenterol, 2005, 11(1):22–26. [7] MEDINA-FRANCO H., HESLIN M. J., CORTES-GONZALES R., Clinicopathological characteristics of gastric carcinoma in young and elderly patients: a comparative study, Ann Surg Oncol, 2000, 7(7):515–519. [8] SU JIN S., BAEK J., KIM H., PARK J., KIM G., KIM D., KIM M., KIM H., MIN Y., Prognostic factors in patients with recurrent gastric cancer after curative gastrectomy, J Clin Oncol, 2008 ASCO Annual Meeting Proceedings (Post-Meeting Edition), 2008, 26(15S):15612.

[9] LI C., KIM S., LAI J. F., OH S. J., HYUNG W. J., CHOI W. H., CHOI S. H., ZHU Z. G., NOH S. H., Risk factors for lymph node metastasis in undifferentiated early gastric cancer, Ann Surg Oncol, 2008, 15(3):764–769. [10] MOGHIMI-DEHKORDI B., SAFAEE A., ZALI M. R., Comparison of colorectal and gastric cancer: Survival and prognostic factors, Saudi J Gastroenterol, 2009, 15(1):18–23. [11] WANG J. Y., HSIEH J. S., HUANG C. J., HUANG Y. S., HUANG T. J., Clinicopathologic study of advanced gastric cancer without serosal invasion in young and old patients, J Surg Oncol, 1996, 63(1):36–40. [12] FUJIMOTO S., TAKAHASHI M., OHKUBO H., MUTOU T., KURE M., MASAOKA H., KOBAYASHI K., Comparative clinicopathologic feature of early gastric cancer in young and older patients, Surgery, 1994, 115(4):516–520. [13] NAOKO T., SHINGO I., YASUKO U., MASATOSHI K., Mixed intestinal- and diffuse-type histology is a risk factor for lymph node metastasis submucosal invasive gastric cancer, Dig Endoscopy, 2008, 20(1):17–21. [14] NOGUCHI Y., Blood vessel invasion in gastric carcinoma, Surgery, 1990, 107(2):140–148. [15] DICKEN B. J., GRAHAM K., HAMILTON S. M., ANDREWS S., LAI R., LISTGARTEN J., JHANGRI G. S., SAUNDERS L. D., DAMARAJU S., CASS C., Lymphovascular invasion is associated with poor survival in gastric cancer: an application of gene-expression and tissue array techniques, Ann Surg, 2006, 243(1):64–73. [16] GABBERT H. E., MEIER S., GERHARZ C. D., HOMMEL G., Incidence and prognostic significance of vascular invasion in 529 gastric-cancer patients, Int J Cancer, 1991, 49(2):203–207. [17] MAEHARA Y., ORITA H., OKUYAMA T., MORIGUCHI S., TSUJITANI S., KORENAGA D., SUGIMACHI K., Predictors of lymph node metastasis in early gastric cancer, Br J Surg, 1992, 79(3):245–247. [18] YOKOTA T., KUNII Y., TESHIMA S., YAMADA Y., SAITO T., TAKAHASHI M., KIKUCHI S., YAMAUCHI H., Significant prognostic factors in patients with node-negative gastric cancer, Int Surg, 1999, 84(4):331–336. [19] HYUNG W. J., LEE J. H., CHOI S. H., MIN J. S., NOH S. H., Prognostic impact of lymphatic and/or blood vessel invasion in patients with node-negative advanced gastric cancer, Ann Surg Oncol, 2002, 9(6):562–567. [20] KOOBY D. A., SURIAWINATA A., KLIMSTRA D. S., BRENNAN M. F., KARPEH M. S., Biologic predictors of survival in node-negative gastric cancer, Ann Surg, 2003, 237(6):828–835; disscussion 835–837. [21] DEL CASAR J. M., CORTE M. D., ALVAREZ A., GARCÍA I., BONGERA M., GONZÁLEZ L. O., GARCÍA-MUÑIZ J. L., ALLENDE M. T., ASTUDILLO A., VIZOSO F. J., Lymphatic and/or blood vessel invasion in gastric cancer: relationship with clinicopathological parameters, biological factors and prognostic significance, J Cancer Res Clin Oncol, 2008, 134(2):153–161. [22] TALAMONTI M. S., KIM S. P., YAO K. A., WAYNE J. D., FEINGLASS J., BENNETT C. L., RAO S., Surgical outcomes of patients with gastric carcinoma: the importance of primary tumor location and microvessel invasion, Surgery, 2003, 134(4):720–727; discussion 727–729.

Corresponding author Daniela Lazăr, Assistant, MD, PhD, Department of Gastroenterology and Hepatology, “Victor Babeş” University of Medicine and Pharmacy, 2 Eftimie Murgu Square, 300041 Timişoara, Romania; Phone +40722–961 955, e-mail: [email protected] Received: March 27th, 2009 Accepted: May 5th, 2009