ANATOMIC PATHOLOGY Original Article
Helicobacter pylori-Related Gastritis and Gastric Ulcer A Continuum of Progressive Epithelial Degeneration KAI M. LEUNG, M.B.B.S., PAK K. HUI, M.D., WING Y. CHAN, M . B . C H B . , AND TERESA M. M. THOMAS, M . B . C H . B .
more than 5 in 25 cases (32.9%) and ED grade more than 5 in 18 cases (23.6%) of GU compared with similar respective scores in 9 cases (17.9%) and 2 cases (3.8%) of ACG. The difference was due primarily to a higher global score of bacterial density and higher focal score of ED in the GU biopsy specimens. These results support the hypothesis that HP-positive ACG and HPpositive GU are lesions within a single disease spectrum. Heavy HP colonization and severe HP-induced epithelial damage are predisposing factors in ulcerogenesis. Because HP-positive ACG is probably a preulcerative state, eradication of the bacteria in HP-positive ACG might prevent subsequent GU. (Key words: Helicobacter pylori; Gastritis; Gastric ulcer; Morphometry) Am J Clin Pathol 1992; 98:569-574
Gastric mucosa colonized by Helicobacter pylori (HP) shows, in addition to acute and chronic inflammation, a specific form of epithelial degeneration (ED),1"8 characterized by loss of apical mucus from the gastric epithelial cells, cell dropout, spotty necrosis, and microerosions.3,5'6 Such changes are absent in HP-negative gastritis and ulcers but are present in almost all HP-positive gastric biopsy specimens in which they are topographically related to the bacteria, regardless of whether the HP-colonized mucosa is ulcerated or simply inflamed.6 Although the etiologic role of HP in active chronic gastritis (ACG) is firmly established,8"15 the relationship between HP and gastric ulcers (GU) is less well defined.4'616 Our previous survey showed that although HP could not be linked to all gastric ulcers, its presence in more than
60% of ulcer biopsy specimens indicated that it may play an important causative role.616 Because the histologic appearances of HP-positive ACG and the ulcer edges of HPpositive GU are remarkably similar, we consider the two lesions different stages of one disease spectrum, with the HP-induced ED playing the key role in ulcerogenesis.7 Therefore, we performed a quantitative morphologic study on a series of HP-colonized gastric mucosal biopsy specimens to determine the relationship between HP colonization, epithelial degeneration, and ulceration. MATERIALS AND METHODS
One hundred forty-five consecutive gastric biopsy specimens showing HP colonization were retrieved from the surgical pathology archives of Kwong Wah Hospital, Hong Kong. The routinely processed paraffin sections were stained with hematoxylin and eosin and a halfGram's stain to enhance the visualization of HP,17 as recFrom the Clinical Pathology Unit, Kwong Wah Hospital, Hong Kong. ommended in the Sydney System for classification of gasReceived December 2, 1991; revised manuscript accepted for publitritis.1819 cation February 10, 1992. Using a system conceptually similar to Gleason's gradAddress reprint requests to Dr. Leung: Clinical Pathology Unit, Kwong Wah Hospital, Waterloo Road, Kowloon, Hong Kong. ing for prostatic carcinoma,20 the degree of HP coloni569
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
One hundred forty-five consecutive gastric biopsy specimens showing colonization by Helicobacter pylori (HP) were studied. Biopsy specimens were obtained from patients with the following conditions: gastric ulcer (GU; 76), active chronic gastritis (ACG; 52), GU with duodenal ulcer (DU; 10), and ACG with DU (7). The mean age of the patients in the ACG group was 8.6 years less than the patients in the GU group. Helicobacter pylori colonization and HP-induced epithelial degeneration (ED) were quantified by a grading system (grades 0 to 6) comprising both focal and global scores for bacterial density (HP grade) and severity of ED (ED grade). The ED grade was directly proportional to the HP grade in all biopsy specimens. Gastric ulcer biopsy specimens were associated with higher HP grades: HP grade
570
ANATOMIC PATHOLOGY Original Article
TABLE 1. SCORING SYSTEM FOR HELICOBACTER PYLORI (HP) COLONIZATION AND EPITHELIAL DEGENERATION (ED) Score1
Definition
Score*
Definition
HP score' = 0 HP score' = 1 HP score' = 2 HP score' = 3 ED score' = 0 ED score' = 1 D score' == 2
No HP At most 10 HP in any gland* 11-30 HP in any gland* >30 HP in any gland* No ED (Fig. \A) Partial loss of mucin (Fig. 2) Full-thickness loss of mucin and/or cell dropout (Fig. 3) Micro- and/or conventional erosions (Fig. 4)
HP score2 = 0 HP score2 = 1 HP score2 = 2 HP score2 = 3 ED score2 = 0 ED score2 = 1 ED score2 = 2
>2 of 3 of glands show HP score' = 0 >1 of 3 of glands show HP score' = 1 >1 of 3 of glands show HP score' = 2 > 1 of 3 of glands show HP score' = 3 >2 of 3 of glands show no ED >1 of 3 of glands show ED score' = 1 >1 of 3 of glands show ED score' = 2
ED score2 = 3
> 1 of 3 of glands show ED score' = 3
ED score' = 3
' Refers to a gland in tubular profile (Fig. I).
RESULTS
FIGS. 1A-D. Scoring system for Helicobacter colonization. (A) HP Score' = 0 (hematoxylin-and-eosin stain; magnification, X250). (B) HP Score' = 1 (hematoxylin-and-eosin stain; magnification, X800). (C) HP Score' = 2. (hematoxylin-and-eosin stain; magnification, X800). (D) HP score' = 3 (hematoxylin-and-eosin stain; magnification X800). The arrows point to the bacteria. Note that in A, the apical cell border is smooth and intact, (i.e.. ED Score' = 0).
The patient data and symptoms are summarized in Table 2. There were 86 cases of GU and 59 cases of ACG. Ten of the GU cases and 7 of the ACG cases also were associated with DU. The mean age of the ACG group was 8.6 years younger than that of the GU group: 40.8% of ACG patients in the present series were younger than 60 years compared with 26% in the GU group (P < 0.01). Pain was a significant symptom in 61.8% of patients with GU, compared with 7.6% of patients with ACG (P < 0.01). The scores and grades for HP colonization and ED are listed in Table 3. Figure 5 graphically represents the relationship between HP grade and ED grade in the patient groups studied. It demonstrates that the severity of epithelial degeneration (ED grade) was, in general, directly proportional to the degree of HP colonization (HP grade).
A.J.C.P. • December 1992
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
zation and ED were each expressed as the sum of two numeric scores. The HP grade was thus HP score1 + HP score2, where HP score1 represented (on a scale of 0 to 3) the degree of bacterial infection at the most heavily colonized focus, and HP score2 indicated the average degree of colonization throughout the biopsy specimen. In a similar fashion, ED grade was expressed as ED score1 + ED score2, where ED score1 represented the degree of epithelial damage at the most severely affected focus, and ED score2 reflected the extent of ED throughout the section. Criteria of the scoring system are summarized in Table 1 and illustrated in Figures 1 to 4. Gastric ulcer was diagnosed by the identification of an ulcer by endoscopic examination and/or the presence of ulcer base necrotic material in histologic sections. Terminologies for ED were as previously described.3,6'7 Duodenal ulcer (DU) was diagnosed by endoscopic examination. Active chronic gastritis was defined by the presence of gastric mucosa showing acute and chronic inflammatory cell infiltration, frequently with intraepithelial neutrophils.2'~23 The chi-squared test was used to compare proportions in different groups.
LEUNG ET AL. Helicobacter pylori-Related Gastritis and Gastric Ulcer
571
FIGS. 4A and B. ED score' = 3. (A) Microerosion, which is a defect, formed as a result of dropout of several epithelial cells. The arrow indicates bacteria in the defect (hematoxylin-and-eosin stain; magnification, X800). (B) Conventional erosion (hematoxylin-and-eosin stain; magnification, XI00).
Although there was overlap between the ACG and GU groups, high HP and ED grades (5 or greater) were found almost exclusively in cases of ulcerated mucosa. All groups showed similar HP score' values (Table 3). None of the ACG cases showed an HP score2 of 3 or an HP grade of 6 (14.5% of the GU group). H. pylori grade of 5 or more was found in 32.9% of cases of GU (17.3% in the ACG group; P < 0.01).
An ED score1 of 3 was shown by 24% of GU cases compared with 7.6% of ACG cases (P < 0.01), whereas an ED grade of 5 or more was shown by 23.6% of the GU group, in contrast to 3.8% of the ACG group (P < 0.01). In the GU cases, the presence or absence of DU was not reflected in any alteration of HP or ED grades. An HP score1 = 3 and HP score2 = 2 or greater was found in more cases of ACG with DU compared with cases of ACG alone. It must be stressed, however, that the sample size was too small for a meaningful statistical analysis. DISCUSSION
FIGS. 3A and B. ED score1 = 2. (A) Complete loss of apical mucin in the foveolar epithelium (arrow head) (hematoxylin-and-eosin stain; magnification, X800). (B) Individual cell dropouts (arrows) (hematoxylinand-eosin stain; magnification, X800).
H. /Ty/on'-colonized gastric mucosa displays characteristic light- and electron-microscopic changes in the epithelium in addition to inflammation.'"7 Salient features are loss of the apical mucous portion of the epithelial cells, resulting in a highly irregular surface with epithelial pits and tufts, dropout of single cells, microerosions, and conventional erosions.3'4'6,7 These epithelial changes are so distinctive that they can serve as histologic criteria to distinguish HP-related forms from other forms of gastritis, such as those associated with reflux or drugs.6,8 Because they occur in biopsy specimens with and without ulcers, we postulate that these progressive epithelial damages are precursor lesions of GU in HP-colonized mucosa.7 The pathogenetic mechanism is uncertain, but urease, lipolytic enzymes, and other toxins produced by HP are possible chemical mediators.24"28 For this study, we designed grading systems to quantify the dose of HP and the severity of ED. The HP and ED
Vol. 98 • No. 6
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
FIGS. 2A and B. ED score1 = 1.(^4) The arrow heads indicate cells showing partial-thickness loss of apical mucin (hematoxylin-and-eosin stain; magnification, X800). (B) The loss of apical mucin is patchy, resulting in pits along the apical cell border (arrow head) (hematoxylin-and-eosin stain; magnification, X800).
572
ANATOMIC PATHOLOGY Original Article TABLE 2. PATIENT DATA AND SYMPTOMS Data and Symptoms
GU
GU + DU
ACG
ACG + DU
Number of cases Male, n (%) Female, n (%) Mean age (years) Age range (years) Age
Helicobacter pylori-Related Gastritis and Gastric Ulcer A Continuum of Progressive Epithelial Degeneration KAI M. LEUNG, M.B.B.S., PAK K. HUI, M.D., WING Y. CHAN, M . B . C H B . , AND TERESA M. M. THOMAS, M . B . C H . B .
more than 5 in 25 cases (32.9%) and ED grade more than 5 in 18 cases (23.6%) of GU compared with similar respective scores in 9 cases (17.9%) and 2 cases (3.8%) of ACG. The difference was due primarily to a higher global score of bacterial density and higher focal score of ED in the GU biopsy specimens. These results support the hypothesis that HP-positive ACG and HPpositive GU are lesions within a single disease spectrum. Heavy HP colonization and severe HP-induced epithelial damage are predisposing factors in ulcerogenesis. Because HP-positive ACG is probably a preulcerative state, eradication of the bacteria in HP-positive ACG might prevent subsequent GU. (Key words: Helicobacter pylori; Gastritis; Gastric ulcer; Morphometry) Am J Clin Pathol 1992; 98:569-574
Gastric mucosa colonized by Helicobacter pylori (HP) shows, in addition to acute and chronic inflammation, a specific form of epithelial degeneration (ED),1"8 characterized by loss of apical mucus from the gastric epithelial cells, cell dropout, spotty necrosis, and microerosions.3,5'6 Such changes are absent in HP-negative gastritis and ulcers but are present in almost all HP-positive gastric biopsy specimens in which they are topographically related to the bacteria, regardless of whether the HP-colonized mucosa is ulcerated or simply inflamed.6 Although the etiologic role of HP in active chronic gastritis (ACG) is firmly established,8"15 the relationship between HP and gastric ulcers (GU) is less well defined.4'616 Our previous survey showed that although HP could not be linked to all gastric ulcers, its presence in more than
60% of ulcer biopsy specimens indicated that it may play an important causative role.616 Because the histologic appearances of HP-positive ACG and the ulcer edges of HPpositive GU are remarkably similar, we consider the two lesions different stages of one disease spectrum, with the HP-induced ED playing the key role in ulcerogenesis.7 Therefore, we performed a quantitative morphologic study on a series of HP-colonized gastric mucosal biopsy specimens to determine the relationship between HP colonization, epithelial degeneration, and ulceration. MATERIALS AND METHODS
One hundred forty-five consecutive gastric biopsy specimens showing HP colonization were retrieved from the surgical pathology archives of Kwong Wah Hospital, Hong Kong. The routinely processed paraffin sections were stained with hematoxylin and eosin and a halfGram's stain to enhance the visualization of HP,17 as recFrom the Clinical Pathology Unit, Kwong Wah Hospital, Hong Kong. ommended in the Sydney System for classification of gasReceived December 2, 1991; revised manuscript accepted for publitritis.1819 cation February 10, 1992. Using a system conceptually similar to Gleason's gradAddress reprint requests to Dr. Leung: Clinical Pathology Unit, Kwong Wah Hospital, Waterloo Road, Kowloon, Hong Kong. ing for prostatic carcinoma,20 the degree of HP coloni569
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
One hundred forty-five consecutive gastric biopsy specimens showing colonization by Helicobacter pylori (HP) were studied. Biopsy specimens were obtained from patients with the following conditions: gastric ulcer (GU; 76), active chronic gastritis (ACG; 52), GU with duodenal ulcer (DU; 10), and ACG with DU (7). The mean age of the patients in the ACG group was 8.6 years less than the patients in the GU group. Helicobacter pylori colonization and HP-induced epithelial degeneration (ED) were quantified by a grading system (grades 0 to 6) comprising both focal and global scores for bacterial density (HP grade) and severity of ED (ED grade). The ED grade was directly proportional to the HP grade in all biopsy specimens. Gastric ulcer biopsy specimens were associated with higher HP grades: HP grade
570
ANATOMIC PATHOLOGY Original Article
TABLE 1. SCORING SYSTEM FOR HELICOBACTER PYLORI (HP) COLONIZATION AND EPITHELIAL DEGENERATION (ED) Score1
Definition
Score*
Definition
HP score' = 0 HP score' = 1 HP score' = 2 HP score' = 3 ED score' = 0 ED score' = 1 D score' == 2
No HP At most 10 HP in any gland* 11-30 HP in any gland* >30 HP in any gland* No ED (Fig. \A) Partial loss of mucin (Fig. 2) Full-thickness loss of mucin and/or cell dropout (Fig. 3) Micro- and/or conventional erosions (Fig. 4)
HP score2 = 0 HP score2 = 1 HP score2 = 2 HP score2 = 3 ED score2 = 0 ED score2 = 1 ED score2 = 2
>2 of 3 of glands show HP score' = 0 >1 of 3 of glands show HP score' = 1 >1 of 3 of glands show HP score' = 2 > 1 of 3 of glands show HP score' = 3 >2 of 3 of glands show no ED >1 of 3 of glands show ED score' = 1 >1 of 3 of glands show ED score' = 2
ED score2 = 3
> 1 of 3 of glands show ED score' = 3
ED score' = 3
' Refers to a gland in tubular profile (Fig. I).
RESULTS
FIGS. 1A-D. Scoring system for Helicobacter colonization. (A) HP Score' = 0 (hematoxylin-and-eosin stain; magnification, X250). (B) HP Score' = 1 (hematoxylin-and-eosin stain; magnification, X800). (C) HP Score' = 2. (hematoxylin-and-eosin stain; magnification, X800). (D) HP score' = 3 (hematoxylin-and-eosin stain; magnification X800). The arrows point to the bacteria. Note that in A, the apical cell border is smooth and intact, (i.e.. ED Score' = 0).
The patient data and symptoms are summarized in Table 2. There were 86 cases of GU and 59 cases of ACG. Ten of the GU cases and 7 of the ACG cases also were associated with DU. The mean age of the ACG group was 8.6 years younger than that of the GU group: 40.8% of ACG patients in the present series were younger than 60 years compared with 26% in the GU group (P < 0.01). Pain was a significant symptom in 61.8% of patients with GU, compared with 7.6% of patients with ACG (P < 0.01). The scores and grades for HP colonization and ED are listed in Table 3. Figure 5 graphically represents the relationship between HP grade and ED grade in the patient groups studied. It demonstrates that the severity of epithelial degeneration (ED grade) was, in general, directly proportional to the degree of HP colonization (HP grade).
A.J.C.P. • December 1992
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
zation and ED were each expressed as the sum of two numeric scores. The HP grade was thus HP score1 + HP score2, where HP score1 represented (on a scale of 0 to 3) the degree of bacterial infection at the most heavily colonized focus, and HP score2 indicated the average degree of colonization throughout the biopsy specimen. In a similar fashion, ED grade was expressed as ED score1 + ED score2, where ED score1 represented the degree of epithelial damage at the most severely affected focus, and ED score2 reflected the extent of ED throughout the section. Criteria of the scoring system are summarized in Table 1 and illustrated in Figures 1 to 4. Gastric ulcer was diagnosed by the identification of an ulcer by endoscopic examination and/or the presence of ulcer base necrotic material in histologic sections. Terminologies for ED were as previously described.3,6'7 Duodenal ulcer (DU) was diagnosed by endoscopic examination. Active chronic gastritis was defined by the presence of gastric mucosa showing acute and chronic inflammatory cell infiltration, frequently with intraepithelial neutrophils.2'~23 The chi-squared test was used to compare proportions in different groups.
LEUNG ET AL. Helicobacter pylori-Related Gastritis and Gastric Ulcer
571
FIGS. 4A and B. ED score' = 3. (A) Microerosion, which is a defect, formed as a result of dropout of several epithelial cells. The arrow indicates bacteria in the defect (hematoxylin-and-eosin stain; magnification, X800). (B) Conventional erosion (hematoxylin-and-eosin stain; magnification, XI00).
Although there was overlap between the ACG and GU groups, high HP and ED grades (5 or greater) were found almost exclusively in cases of ulcerated mucosa. All groups showed similar HP score' values (Table 3). None of the ACG cases showed an HP score2 of 3 or an HP grade of 6 (14.5% of the GU group). H. pylori grade of 5 or more was found in 32.9% of cases of GU (17.3% in the ACG group; P < 0.01).
An ED score1 of 3 was shown by 24% of GU cases compared with 7.6% of ACG cases (P < 0.01), whereas an ED grade of 5 or more was shown by 23.6% of the GU group, in contrast to 3.8% of the ACG group (P < 0.01). In the GU cases, the presence or absence of DU was not reflected in any alteration of HP or ED grades. An HP score1 = 3 and HP score2 = 2 or greater was found in more cases of ACG with DU compared with cases of ACG alone. It must be stressed, however, that the sample size was too small for a meaningful statistical analysis. DISCUSSION
FIGS. 3A and B. ED score1 = 2. (A) Complete loss of apical mucin in the foveolar epithelium (arrow head) (hematoxylin-and-eosin stain; magnification, X800). (B) Individual cell dropouts (arrows) (hematoxylinand-eosin stain; magnification, X800).
H. /Ty/on'-colonized gastric mucosa displays characteristic light- and electron-microscopic changes in the epithelium in addition to inflammation.'"7 Salient features are loss of the apical mucous portion of the epithelial cells, resulting in a highly irregular surface with epithelial pits and tufts, dropout of single cells, microerosions, and conventional erosions.3'4'6,7 These epithelial changes are so distinctive that they can serve as histologic criteria to distinguish HP-related forms from other forms of gastritis, such as those associated with reflux or drugs.6,8 Because they occur in biopsy specimens with and without ulcers, we postulate that these progressive epithelial damages are precursor lesions of GU in HP-colonized mucosa.7 The pathogenetic mechanism is uncertain, but urease, lipolytic enzymes, and other toxins produced by HP are possible chemical mediators.24"28 For this study, we designed grading systems to quantify the dose of HP and the severity of ED. The HP and ED
Vol. 98 • No. 6
Downloaded from http://ajcp.oxfordjournals.org/ by guest on June 6, 2016
FIGS. 2A and B. ED score1 = 1.(^4) The arrow heads indicate cells showing partial-thickness loss of apical mucin (hematoxylin-and-eosin stain; magnification, X800). (B) The loss of apical mucin is patchy, resulting in pits along the apical cell border (arrow head) (hematoxylin-and-eosin stain; magnification, X800).
572
ANATOMIC PATHOLOGY Original Article TABLE 2. PATIENT DATA AND SYMPTOMS Data and Symptoms
GU
GU + DU
ACG
ACG + DU
Number of cases Male, n (%) Female, n (%) Mean age (years) Age range (years) Age
