Association between cag-pathogenicity island in Helicobacter pylori isolates from peptic ulcer, gastric carcinoma, and non-ulcer dyspepsia subjects with histological changes

Abstract

AIM: To investigate the presence of the cag-pathogenicity island and the associated histological damage caused by strains with complete cag-PAI and with partial deletions in correlation to the disease status.

METHODS: We analyzed the complete cag-PAI of 174 representative Helicobacter pylori (H pylori ) clinical isolates obtained from patients with duodenal ulcer, gastric ulcer, gastric cancer, and non-ulcer dyspepsia using eight different oligonucleotide primers viz cagA1, cagA2, cagAP1, cagAP2, cagE, cagT, LEC-1, LEC-2 spanning five different loci of the whole cag-PAI by polymerase chain reaction (PCR).

RESULTS: The complete screening of the genes comprising the cag-PAI showed that larger proportions of subjects with gastric ulcer (97.8%) inhabited strains with complete cag-PAI, followed by gastric cancer (85.7%), non-ulcer dyspepsia (7.1%), and duodenal ulcer (6.9%), significant differences were found in the percentage distribution of the genes in all the clinical groups studied. It was found that strains with complete cag-PAI were able to cause severe histological damage than with the partially deleted ones.

CONCLUSION: The cag-PAI is a strong virulent marker in the disease pathogenesis as it is shown that a large number of those infected with strain with complete cag-PAI had one or the other of the irreversible gastric pathologies and interestingly 18.5% of them developed gastric carcinoma. The presence of an intact cag-PAI correlates with the development of more severe pathology, and such strains were found more frequently in patients with severe gastroduodenal disease. Partial deletions of the cag-PAI appear to be sufficient to render the organism less pathogenic.

Key Words: Helicobacter pylori, cag-pathogenicity island, Genetic diversity, Gastro-duodenal diseases

INTRODUCTION

Gastric cancer is the second most deadly malignant neoplasia worldwide. According to the presently available statistics, approximately 74% of those diagnosed succumb to this disease every year[1]; this is because of poor prognosis as it is often made when the disease has assaulted the muscularis propria. Evidences show that the pathogenesis of gastric cancer is a multistep process[2,3]. This ‘cascade’ is believed to be triggered by Helicobacter pylori (H pylori ) infection, a Gram negative pathogen. Chronic infection with this gastric pathogen is known to be the major factor driving the precancerous process via mechanisms including direct transformation of cells, induction of immunosuppression with consequently reduced cancer immunosurveillance, or by causing chronic inflammation[4,5]. In 1994, the International Agency for Research on Cancer (IARC) declared H pylori a Class l (definite) carcinogen based on the epidemiological and interventional studies in human beings[6] and convinced that this bacterial infection indeed plays a key role in the initiation of the neoplastic process in the stomach.

Although many attempts in the past have been made to understand and associate the causal link between H pylori infection and the sequelae that leads to gastric carcinoma[7,8], there are conflicting data in the literature due to differences in the study population and designs[9,10].

In the past few years, many H pylori virulence factors have been described that contribute to the survival of this pathogen in an extremely hostile acidic milieu of the stomach and its colonization in that organ[11,12]. Other putative virulence determinants such as vacuolating cytotoxin gene A (vacA), cytotoxin associated gene pathogenicity island (cag-PAI) and induced by contact with epithelium gene A (iceA) are not present in all H pylori strains or are known to exhibit different allelic variations[13]. The cytotoxin-associated gene island also referred to as the cag-PAI is an approximately 40-kb cluster of genes and is the most studied marker of the H pylori. In many studies[20,22,24], this large fragment has been a criterion of typing H pylori into pathogenic and non-pathogenic strains (Type I and Type II). Studies emphasizing on the functional importance of this island have reported that strains possessing cag-PAI induce more notable phenotypic changes in vivo, such as higher levels of IL-8 production than cag-PAI negative ones[30]. Recent studies made from within our institute[14,15] and other parts of the world[16,17] have mainly focused on the presence of this large fragment and its association with the disease status. Therefore, in a continuing attempt to further establish a strong epidemiologic relation between the cag-PAI and the disease conditions with reference to the histopathologic changes lead to the inception of the present study. As it is reported[18] that the presence of cag island and the consequent cag instability may produce differences in the pathogenicity and host adaptability within a bacterial strain, detailed analyses of the genes of the cag island in human beings isolates with reference to histologic damage and disease outcome would be essential. Therefore, compelled by these observations, the present study was designed to identify the distribution of different genes of the cag-PAI in clinical H pylori isolates by assessing the presence of representative genes located in different segments of the cag-island and its correlation with the histologic changes and disease status.

MATERIALS AND METHODS

Patients and sampling

The study population consisted of 174 patients (100 males and 74 females) with a mean age of 48.4 years (range 21-73 years). The patients were classified at the time of endoscopy into those suffering from active ulcers disease [duodenal ulcers (n = 58), gastric ulcers (n = 46), gastric carcinoma (n = 14)] and those with no evidence of mucosal ulcer and gastritis, but suffering from mild or severe dyspeptic symptoms, i.e. non-ulcer dyspepsia (n = 56). None of the patients included in the study had received NSAIDs or antibiotics within the previous 2 mo. Informed consents were taken from the patients who underwent upper gastrointestinal endoscopy at the department of gastroenterology, Deccan College of Medical Sciences, Hyderabad.

Four gastric biopsies were collected: one in urea solution for the rapid urease test (RUT), one in supplemented broth for isolating culture, one in phosphate-buffered saline (PBS) for testing by PCR assay, and one in 10% buffered formalin for histological examination by modified Giemsa stain for the presence of H pylori.

H pylori strains

A total of 174 clinical H pylori strains were screened for the presence of cag-PAI genes. These strains were recovered from individual subjects undergoing upper gastrointestinal endoscopy presenting with various symptoms. This included 43 live strains and 131 genomic DNA isolated from the gastric biopsy.

Extraction of genomic DNA

H pylori culture and DNA extraction from the culture and biopsy was carried out as described elsewhere[19].

PCR analysis of the cag-PAI genes

The genes of the cag-PAI were PCR amplified under the conditions described by Ikenoue et al[20]. One microliter of the extracted genomic DNA was used in a 20 μL reaction volume containing 1´ PCR buffer, 1 U Taq DNA polymerase, 1.5 mmol/L Mg2⁺, 200 μmol/L each dNTP and 10 pmol/L of each primer.

The cycling parameters were optimized and are as follows: Initial denaturation at 95 °C for 5 min, followed by 40 cycles each of denaturation at 94 °C for 30 s, annealing at 52 °C for 30 s and extension at 72 °C for 1.5 min and finally after the last cycle, extension was continued for another 7 min.

Eight sets of oligonucleotide primers spanning the 40 kb cag-PAI were used in the study. Appropriate positive and negative controls were included in each set to avoid misinterpretation of results. The details of the primers used with their product sizes are enlisted in Table 1. Amplicons were separated by electrophoresis on 2% agarose gel and stained by ethidium bromide.

Table 1 List of primers.

Target	Primer	Sequence (5’- 3’)	Product
gene			size (bp)
	CagA-F1	AACAGGACAAGTAGCTAGCC
CagA1	CagA-R1	TATTAATGCGTGTGTGGCTG	701
	CagA-F2	GATAACAGGCAAGCTTTTGA
CagA2	CagA-R2	CTGCAAAAGATTGTTTGGCAGA	349
	CagAP-F1	GTGGGTAAAAATGTGAATCG
CagAP1	CagA-R1	TATTAATGCGTGTGTGGCTG	730
	CagAP-F2	CTACTTGTCCCAACCATTTT
CagAP2	CagA-R2	CTGCAAAAGATTGTTTGGCAGA	1 181
	CagE-F1	GCGATTGTTATTGTGCTTGTAG
CagE	CagE-R1	GAAGTGGTTAAAAAATCAATGCCCC	329
	CagT-F1	CCATGTTTATACGCCTGTGT
CagT	CagT-R1	CATCACCACACCCTTTTGAT	301
	LEC-F1	ACATTTTGGCTAAATAAACGCTG
LEC-1	LEC-R1	TCTCCATGTTGCCATTATGCT	384
	LEC-F2	ATAGCGTTTTGTGCATAGAA
LEC-2	LEC-R2	ATCTTTAGTCTCTTTAGCTT	877

Histopathological analysis

The biopsy specimen collected from the gastric antrum was used for histopathologic examination to grade the severity of disease after they were embedded in paraffin and stained with hematoxylin and eosin. A single experienced pathologist (ZA), who was blinded to the patient’s history and molecular data of the isolate, evaluated all histological data. The grade of gastritis was determined on the basis of the updated Sydney system[21].

Statistical analysis

The data were analyzed by means of χ² test and the Mann–Whitney U test.

RESULTS

In a total of 174 isolates screened, only 65 (37.4%) was found to carry the complete cag-PAI, while 109 (62.6%) carried the cag-PAI with partial deletions. No isolate was found with completely deleted cag-PAI (Table 2).

Table 2 Relationship between the presence of cag-PAI and the clinical status.

Clinical status (n)	Intact PAI (%)	cag-PAI type	Completely deleted
		Partially deleted	PAI (%)
		PAI (%)
DU (58)	4 (6.9)	54 (93.1)	0
GU (46)	45 (97.8)	1 (2.2)	0
NUD (56)	4 (7.1)	52 (92.9)	0
GC (14)	12 (85.7)	2 (14.3)	0
Total (174)	65 (37.4)	109 (62.6)	0

χ²=130.71 .

With reference to the clinical status, we found majority of the duodenal ulcer (DU) isolates and non-ulcer dyspepsia (NUD) isolates to possess partially deleted cag-PAI, while on the contrary we found ~97.8% of the gastric ulcer (GU) isolates and 85.7% of the gastric carcinoma (GC) isolates possessed the intact cag-PAI. The details of the results obtained are given in Table 2 and Figure 1. Statistically these differences were highly significant.

Figure 1 Distribution of cag-PAI in dyspeptic subjects.

When checked for the presence of each locus, we could find that the total cagA gene (i.e. with promoters) to be present in 69 (39.7%) of all the isolates screened, 4 (2.3%) isolates had total deletion of the gene and 101 (58%) were carrying partial deletions. Of the 69-cagA positive isolates, prevalence of cagA gene was predominant among gastric ulcer (GU) and gastric carcinoma (GC) strains (Table 3 and Figure 2) simultaneously cagA gene with partial deletions were more prevalent among H pylori isolated from duodenal ulcer (DU) and non-ulcer dyspepsia (NUD) subjects.

Figure 2 Distribution of cagA in clinical isolates.

Table 3 cagA status of H pylori isolates.

Clinical status (n)	cagA +ve (%)	Partial cagA +ve (%)	cagA–ve (%)
DU (58)	8 (13.8)	46 (79.3)	4 (6.9)
GU (46)	45 (97.8)	1 (2.2)	0
NUD (56)	4 (7.1)	52 (92.9)	0
GC (14)	12 (85.7)	2 (14.3)	0
Total (174)	69 (39.7)	101 (58)	4 (2.3)

DU vs GU, DU vs GC, GU vs NUD and NUD vs GC combinations are highly significant at 0.1% level. DU vs NUD and GU vs GC combinations are not significant at 0.1% level.

Observing the prevalence patterns of cagA deletion mutant strains, we further analyzed the predominance of deletions in the promoter and body part of the cagA gene. The body part of cagA gene (i.e. A1+A2) was present in a maximum number of isolates i.e. 134 (77%), among which a considerable number of isolates lacked the promoter region (i.e. AP1+AP2) (Table 4).

Table 4 Distribution of cagA gene and cagA promoter in H pylori isolates.

Clinical status (n)	A1+A2 (%)	AP1+AP2 (%)
DU (58)	43 (74.1)	8 (13.8)
GU (46)	45 (97.8)	46 (100)
NUD (56)	32 (57.1)	6 (10.7)
GC (14)	14 (100)	12 (85.7)
Total (174)	134 (77)	72 (41.4)

Screening of the cagII region of the cag-PAI revealed cagE, cagT, LEC-1, and LEC-2 to be present in 143 (82.1%), 145 (87.3%), 142 (81.6%) and 111 (63.7%) of the total isolates, respectively. The distribution of these genes with reference to disease status is illustrated in detail in Table 5 and Figure 3. We found the difference between DU and GU, GU and NUD, GC and NUD to be highly significant statistically.

Figure 3 Distribution of cagE, T, LEC1, LEC2

Table 5 Distribution of cagE, T, LEC1, and LEC2 in H pylori isolates.

Clinical status (n)	cagE (%)	cagT (%)	LEC1 (%)	LEC2 (%)
DU (58)	47 (81)	56 (96.5)	44 (75.8)	33 (56.8)
GU (46)	46 (100)	46 (100)	46 (100)	46 (100)
NUD (56)	36 (64.2)	36 (64.2)	38 (67.8)	18 (32.1)
GC (14)	14 (100)	14 (100)	14 (100)	14 (100)
Total (174)	143 (82.1)	152 (87.3)	142 (81.6)	111 (63.7)

For cagE, GU vs NUD is highly significant at 0.1% level (P<0.001), DU vs GU, NUD vs GC combinations are significant at 1% level (P<0.01), DU vs NUD combination is significant at 5% level (P<0.05) and DU vs GC . For cagT, DU vs NUD, GU vs NUD combinations are highly significant at 0.1% level (P<0.001), NUD vs GC combination is significant at 1% level (P<0.01) and DU vs GU, DU vs GC, GU vs GC are not significant at 1% level. For LEC1, DU vs GU, GU vs NUD combinations are highly significant at 0.1% level (P<0.001), NUD vs GC is significant at 1% level (P<0.01), DU vs GC is significant at 5% level (P<0.05). For LEC2, DU vs GU, GU vs NUD, NUD vs GC combinations are highly significant at 0.1% level (P<0.001), DU vs NUD, DU vs GC combinations are significant at 1% level (P<0.01) and GU vs GC is not significant at 1% level.

In the present study, histopathological examination of antral biopsies from a total of 174 subjects was carried out to check the grade of gastritis and the differences between the pathologies of intact cag-PAI and partial cag-PAI infection. Of the 174 subjects included for the study, we screened 160 subjects as 14 of them had endoscopically proven gastric carcinoma and hence histopathology confirmed features of carcinoma. Out of the 160 subjects screened, all showed chronic gastritis and on further screening for any other advanced type of gastritis, we found 112 (70%) of the total 160 to possess topographic chronic superficial gastritis, 20 (12.5%) showed atrophic changes, whereas 18 (11.25%) and 10 (6.25%) of the subjects showed IM and dysplasia respectively (Table 6). When each of these histological lesions were scored individually in relation to their respective clinical status, we found that among the 58 duodenal ulcer cases, 52 (89.7%) showed chronic superficial gastritis, 2 (3.4%) showed atrophy and 4 (6.9%) showed intestinal metaplasia while none of the subjects had dysplasia (Table 6).

Table 6 Frequency of atrophy, metaplasia, and dysplasia in chronic gastritis subjects.

Clinical	Acute	Chronic gastritis
status (n)	gastritis (%)	Superficial	Gastritis with	Gastritis with	Gastritis with
		gastritis	atrophy (%)	metaplasia (%)	dysplasia (%)
DU (58)	0	52 (89.7)	2 (3.4)	4 (6.9)	0
GU (46)	0	9 (19.6)	13 (28.3)	14 (30.4)	10 (21.7)
NUD (56)	0	51 (91.1)	5 (8.9)	0	0
Total (160)	0	112 (70)	20 (12.5)	18 (11.25)	10 (6.25)

For atrophy, DU vs GU combination is highly significant at 0.1% level (P<0.001), GU vs NUD is significant at 1% level (P<0.01) and DU vs NUD combination is not significant at 1% level. For metaplasia, DU vs GU, GU vs NUD combinations are highly significant at 0.1% level (P<0.001) and DU vs NUD combination is significant at 5% level (P<0.05). For dysplasia, DU vs GU, GU vs NUD combinations are highly significant at 0.1% level (P<0.001) and DU vs NUD combination is not significant at any level.

In the gastric ulcer category, we found that 9 (19.6%) showed chronic superficial gastritis, 13 (28.3%) showed atrophy, 14 (30.4%) showed intestinal metaplasia and 10 (21.7%) showed dysplasia. For atrophy, metaplasia and dysplasia, the difference between GU and DU was statistically highly significant (P<0.001, Table 6).

In the NUD category, we found that 51 (91.1%) showed chronic superficial gastritis and 5 (8.9%) showed atrophic gastritis while none had intestinal metaplasia and dysplasia. For metaplasia and dysplasia, the difference between GU and NUD, was statistically highly significant (P <0.001, Table 6).

When we analyzed the histological status, with reference to the strain infecting with either intact PAI or partially deleted ones, we found a significant difference among them. As evidenced from Table 7, we found that, of the 65 subjects infected with intact PAI, 8 (12.3%) showed chronic superficial gastritis, 17 (26.1%) showed atrophic changes, 18 (27.7%) showed intestinal metaplasia (all of them were Type III), 10 (15.4%) showed high grade dysplastic changes and 12 (18.5%) showed intestinal type gastric carcinoma. On the contrary, when we checked those subjects, who inhabited partially deleted strains, we found 104 (95.4%) to possess chronic superficial gastritis, 3 (2.8%) showed atrophy while 2 (1.8%) showed gastric carcinoma and these differences were statistically very significant (P<0.001, Table 7).

Table 7 Histological status of subjects with intact cag-PAI (n = 65) and partial cag-PAI (n = 109).

Cag-type	Chronic gastritis							Gastric carcinoma
(n)	Superficial gastritis	Gastritis with atrophy	Gastritis with metaplasia (%)			Gastritis with dysplasia(%)		Intestinal (%)	Diffuse (%)
	(%)	(%)	Ty-1	Ty-2	Ty-3	Low grade	High grade
Intact (65)	8 (12.3)	17 (26.1)b	0	0	18	0	10	12b	0
					-27.7		-15.4	-18.5
Partial (109)	104 (95.4)	3 (2.8)	0	0	0	0	0	2 (1.8)	0

^bP<0. 001 vs others.

When a correlation between the histological status of the subjects infected with either intact cag-PAI or partially deleted strains isolated from subjects with varied disease status was made, we found that among 58 DU subjects, 4 had intact cag-PAI and all of them were Type III intestinal metaplasia, among the remaining 54, which had partial deletions in the cag-PAI, 52 (96.3%) showed chronic superficial gastritis and 2 (3.7%) showed atrophic changes. Among the GU subjects, of the 46 infected, 45 were known to possess the cag-PAI and 8 (17.8%) of them showed chronic superficial gastritis, 13 (28.9%) showed atrophic gastritis, 14 (31.1%) showed Type III intestinal metaplasia and 10 (22.2%) showed high grade dysplastic changes (Table 8). Among the 56 NUD subjects, 4 with intact cag-PAI showed atrophic changes and of the rest with partial deletions 51 (98.1%) showed chronic superficial gastritis and 1 (1.9%) showed atrophic gastritis (Table 8) while among the 14 isolated from gastric carcinoma 12 had intact cag-PAI while the other 2 partial deletions and when scored for the type of gastric carcinoma all of them showed intestinal type of carcinoma (Table 8).

Table 8 Correlation of the histological status of the subjects with intact and partially deleted cag-PAI from varied disease status.

Clinical status	Chronic gastritis								Gastric carcinoma
(n)	cag status	Superficial gastritis	Gastritis with atrophy (%)	Gastritis with metaplasia (%)			Gastritis with dysplasia (%)		Intestinal (%)	Diffuse (%)
	(n)	(%)		Ty-1	Ty-2	Ty-3	Low grade	High grade
DU (58)	Intact (4)	0	0	0	0	4 (100)	0	0	0	0
	Partial (54)	52 (96.3)	2 (3.7)	0	0	0	0	0	0	0
GU (46)	Intact (45)	8(17.8)a	13 (28.9)	0	0	14 (31.1)	0	10 (22.2)	0	0
	Partial (1)	1 (100)	0	0	0	0	0	0	0	0
NUD (56)	Intact (4)	0	4(100)a	0	0	0	0	0	0	0
	Partial (52)	51 (98.1)	1 (1.9)	0	0	0	0	0	0	0
GC (14)	Intact (12)	0	0	0	0	0	0	0	12 (100)b	0
	Partial (2)	0	0	0	0	0	0	0	2 (100)	0

^aP <0.05, ^bP <0.001 vs others.

DISCUSSION

The cag-PAI is an approximately 40-kb cluster of genes in H pylori chromosome, and a quite conservative entity. Censini et al[22] in 1996 first identified strains with partially deleted cag-PAIs. The molecular mechanism of these genetic rearrangements was explained by incorporation of an insertion element, IS605, in cag-PAI. Recently, the composition of the cag-PAI in clinical H pylori isolates has been studied in different populations by various methods, including PCR, dot blotting and long distance PCR[20,23,24].

In the present study, we used simple PCR for structural screening of cag-PAI in clinical isolates of H pylori. Out of the 174 clinical isolates, we found only 37.4% were carrying the complete cag-PAI (Table 2), whereas Mukhopadhyay et al[25] reported more than 96% in Calcutta strains of peptic ulcer and non-ulcer dyspepsia. In our study, 97.8% of gastric ulcer and 85.7% (Table 2 and Figure 1) of the gastric carcinoma strains were carrying complete PAI, which are considered to be severe forms of the gastro-duodenal diseases. Even though duodenal ulcer is also considered to be a severe form of the gastro-duodenal disease, the proportion of DU strains that carried was just 6.9%. Jenks et al[23] reported that the presence of certain genes (cagA, cagE, cagM, T, ORF 6, 10, 13) in the cag-PAI is highly associated with duodenal ulcers. We too found a similar type of observation, but not for all the seven genes which they selected. We observed the same kind of correlation with only two genes i.e. cagE and cagT, where DU cases carried cagE with 81% and cagT with 96.5% (Table 5 and Figure 3). Whereas NUD isolates were carrying the genes with an average of 65% (Table 5 and Figure 3). Further, Day et al[26] revealed that isolates containing cagE were associated with duodenal ulceration.

Among non-ulcer dyspepsia strains, we found 7.1% to carry the complete cag-PAI, which is statistically almost equal to DU percentage, and a report from Sweden[27] showed 58% of cag-PAI positivity in NUD isolates. In the same report, the authors showed that 5% of isolates from severe pathology i.e. gastric carcinoma and duodenal ulcer, and 15% of the isolates from NUD lacking the cag-PAI. Not only the data from other continents, but from the same Indian sub-continent showed total deletions from ulcer group and non-ulcer group[25], whereas we could not come across a single isolate with entirely deleted cag-PAI from any of the disease condition indicating strain diversity. This kind of diversity, i.e. absence of completely deleted PAIs and presence of just 6.9% complete cag-PAIs in duodenal ulcer cases, might be particularly true for the south Indian of Telugu linguistic group who are mainly Dravidian and married consanguineously for millennia[28]. Their genetic separation from other Indian communities during much of the human history has already been reported[29].

Strains with intermediate genotypes, lacking parts of the cag-PAI, were found in 62.6% (Table 2) and more frequently found in patients with non-ulcer dyspepsia and duodenal ulcer (Figure 1). A probable mechanism for the establishment of these internal deletions within the cag-PAI would be that the short repeated sequences found by Nilsson et al[27] may serve as homologies enabling slipped strand mispairing and consequently excision of the enclosed DNA fragment especially in DU and NUD subjects. Moreover, in our observation, the partially deleted cag-PAI represented a genotype more common than a complete cag-PAI and no strain was found with completely deleted cag-PAI.

Conventionally cagA was used and is still used as a marker for the presence of an intact cag-PAI and for virulence. Recently, Backert et al[30] showed a good correlation of cagA with the presence of cag-PAI. In this study, we found that 39.7% of the strains carried the complete cagA gene (Table 3) and the presence of cagA gene did not correlate with the genetic presence of complete cag-PAI. Further, for - 4 kb cagA gene, we used four sets of primers i.e. two primers for body of the gene and two primers for promoter, designed from various locations of the body region and promoter, whereas other studies taken up earlier had used a single set of primer for complete cag-PAI. This might be the reason for the correlation, which they obtained. The typical observation was when 77% of the isolates were carrying the body region (A1+A2) only 41.4% of them carried promoter region (Table 4), which means that in our isolates even though the strains carried the cagA gene, most of them lacked the promoter of the gene, without which cagA is not functional. Further, four isolates i.e. 2.3% completely lacked the cagA gene and all the four isolates belonged to duodenal ulcer cases.

Among many virulence markers present in the H pylori genome, cag-PAI is the major virulence factor and is associated with severe gastroduodenal pathology[31,32] that includes both duodenal and gastric ulcers along with carcinomas. Some studies have identified a correlation between an intact cag-PAI and development of disease[20,23,24,27], as we are trying to show with this present study in Indian scenario, whereas others could not find such a relationship[33,34].

“Infection with H pylori always causes chronic active gastritis”[35]. This phrase has become true in our observation. As it is observed in Table 6, there are no acute gastritis subjects. Moreover, the subjects infected with cag-PAI positive strains were found to show severe forms of histopathological changes, like atrophic gastritis, intestinal metaplasia, and neoplasia. This might be the reason for IARC-WHO to designate H pylori a class I (definite) carcinogen[6].

Out of the 174 isolates, 65 (37.4%) had complete cag-PAI and 109 (62.6%) had partial deletions in the cag-PAI (Table 2). As evident from Table 7, it can be observed that subjects infected with H pylori strains with intact cag-PAI had many remarkable histopathological changes, when compared to those who had partially deleted cag-PAI. It is quite clear from the statistics (P<0.001) that among 65 H pylori strains with intact cag-PAI, 18.5% subjects had advanced cancerous lesions, while only 1.8% of the 109 subjects, who harbored H pylori with partial cag-PAI had advanced to carcinoma thus allowing us to delineate that persons with intact cag-PAI are 10-fold more prone to develop carcinoma in comparison to the partially deleted cag-PAI strains.

Parallelly, a small group of population, i.e. 12.3% with intact cag-PAI, were shown to have only chronic superficial gastritis, but according to Ohkuma et al[36] H pylori positive cases with chronic gastritis have increased risk of atrophy and intestinal metaplasia. On the other hand, partial cag-PAI subjects were also shown to have chronic gastritis (Table 6), but the percentage of disease progression was very high among cag-PAI positive subjects (Table 7) than those with partially deleted ones. Moreover, Type-3 metaplasia and high-grade dysplasia were seen only in cag-PAI positive subjects, where Type-3 metaplasia is closely linked to carcinoma[37] and dysplasia is nothing but a non-invasive type of neoplasia[38]. The results of this study are in contrast to those obtained by Keates et al[39] who determined that gene products of the cag pathogenicity island are required for maximal activation of mitogen-activated protein kinases (MAPK) in gastric epithelial cells, which regulate cell proliferation, differentiation, inflammatory responses, stress, and programmed death, leading to induce gastroduodenal inflammation, ulceration and neoplasia. Further, Naumann et al[40] stated that the integrity of whole cag-PAI is also a pre-requisite for efficient activation of early transcription factor AP-1, which is known for its immuno-stimulatory function.

In relation with disease status, among DU, GU, and NUD, gastric ulcers are considered to be more prone to the gastric carcinoma[41]. The observations of this study are in correlation to that obtained by Hansson et al[41]. When we compare GU subjects with DU and NUD, the percentage of predisposing factors was much more among GU subjects. Recently, Wanatabe et al[42] proved this in animal models. In their study at 26 wk, Mongolian gerbils developed chronic gastritis, ulceration and metaplasia. At 62 wk, 31% of them developed adenocarcinoma. Interestingly the inoculum used for the infection was obtained from a patient with gastric ulcer. Moreover, there have been reports that gastric cancer mortality rates bear an inverse relationship to duodenal ulcer disease rates[43], suggesting that they are directly relating with gastric ulcer in ulcer groups.

Further, in partial cag-PAI subjects, 2.8% showed atrophy, 1.8% showed carcinoma (Table 7) suggesting the role of other virulence genes and risk factors. Parallelly high incidence rate of gastric carcinoma among the gastric ulcer cases might be true, but it should not be assigned to a single determinant such as cag-PAI, but it is a result of many factors such as host genetic factors, environment, low socio-economic status, irregular dietary habits in addition to H. pylori with complete cag-PAI[43].

Hence, we can suggest that the cag-PAI is a strong virulent marker in the disease pathogenesis because more than 85% of the cag-PAI positive subjects were shown to have one or the other of the irreversible gastric pathologies and interestingly 18.5% of them developed gastric carcinoma and GU is the major risk/predisposing factor for the gastric carcinoma. Moreover, duodenal ulcer is not at all a risk factor for severe gastric pathologies and it is not a severe kind of disease, like gastric ulcer in our population.