Primary gastric lymphoma of MALT : Considerations of pathogenesis , diagnosis and therapy

About 40% of all non-Hodgkin’s lymphoma arise from an extranodal origin. Although many specifics in its clinical presentation and biological behaviour have become evident, extranodal lymphoma in general, and primary gastrointestinal lymphoma in particular, have been typed according to classification systems established for nodal lymphoma (Rappaport, Kiel, Working Formulation, etc) for decades. This has made comparing clinical data difficult or

even impossible, particularly because the definition of primary gastrointestinal lymphoma was not uniform.

DEFINITION
All prior definition criteria were exclusively based on the pattern of dissemination of the disease (1)(2)(3).Due to the outstanding work of Isaacson and Spencer (4) in the 1980s and the establishment of the concept of mucosa-associated lymphoid tissue (MALT), primary gastrointestinal lymphoma is now considered a distinct entity with its own histological classification (5) (Table 1).The vast majority of primary gastric lymphoma is B cell lymphoma of MALT (groups 1 and 2 of column 1 in Table 1), which can be identified by means of histopathological, immunohistochemical and cellular biological characteristics.Only in cases of high grade lymphoma without low grade components does the dissemination pattern have to be considered to separate widespread nodal disease, which secondarily involves the gastrointestinal tract (6).The revised European American classification of lymphoid neoplasms (REAL classification), which was proposed in 1994 (7) and is gaining importance, also lists B cell lymphoma of MALT as a distinct group (named extranodal marginal-zone B cell lymphoma).Now, for the first time, comparative analysis of international studies is feasible.

PATHOGENESIS
Gastrointestinal lymphoma is by far the most frequent manifestation of all extranodal lymphomas.Within the digestive tract, gastric lymphoma accounts for 70% or more of all lymphomas.This is surprising given that, contrary to the physio-logical MALT of the small and large bowel, no lymphatic tissue occurs in the normal mucosa.This finding raises questions about the origin of acquired MALT and potential preneoplastic conditions.In the late 1980s, acquisition of intramucosal lymph follicles and accumulation of immunoglobulin A-producing plasma cells were discovered to result from Helicobacter pylori infection of the gastric mucosa (8)(9)(10).This acquired lymphatic tissue showed morphological characteristics of MALT and regressed after successful eradication of the bacterium (11).In 1991, Wotherspoon et al (12) demonstrated for the first time that patients with primary gastric MALT lymphoma are regularly infected by Can J Gastroenterol Vol 14 Suppl D November 2000 45D Gastric MALT lymphoma H pylori.This finding has been confirmed by other investigators (13,14).In addition to histomorphological studies, recent epidemiological (15,16), molecular biological (17)(18)(19) and experimental (20,21) data clearly indicate that H pylori plays a decisive role in the development and progression of gastric MALT lymphoma.Figure 1 summarizes current concepts of the pathogenesis of this lymphoma.The initial antigen-dependent proliferation is of major importance for an antibiotic treatment approach, as is discussed later.As long as antigen-driven tumour proliferation is evident, successful elimination of this stimulus may be followed by regression of the lymphoma.There is a need, however, to identify the parameters for the progression from H pylori-dependent lymphoma to autonomous tumour growth.The recent finding of t (11;18) (q21;q21) chromosome translocation may be a preliminary step toward this effort (22).It may be assumed that high grade transformation does not always follow the sequence described (Figure 1).The frequency of de novo blastic, high grade lymphoma is still unknown.Also, other microorganisms may be of potential pathogenetic importance.Stolte et al (23) recently reported on a possible association of Helicobacter heilmannii with gastric MALT lymphoma.This consideration is also supported by findings of lymphoma regression in some serologically and histologically H pylori-negative individuals following usual anti-H pylori, anti-infectious therapy (24).

DIAGNOSIS AND CLINICAL STAGING PROCEDURES
The first two studies, which based their retrospective analysis of large populations on the MALT classification, clearly identified malignancy (low grade versus high grade) and stage as the two major prognostic factors and therapeutic determinants (25,26).Endoscopic biopsies and endoscopic ultrasounds are, therefore, of particular importance to diagnose, classify and stage gastric lymphoma reliably, especially if primary conservative therapy is favoured and pathohistological evaluation of the resected specimen is not available.Potential problems of endoscopic bioptic diagnoses are based on: • the unspecific macroscopic appearance of the lymphoma ( 27); • a submucosal growth pattern without visible changes; • the multilocality of the lymphoma; • the possibility of high malignant transformation arising focally; and • a sampling error due to few and/or not representative specimens.
These theoretical considerations may be confirmed by the results of a study that compared the accuracy of endoscopic biopsies with the definite diagnosis of the gastrectomy specimens by the same reference pathologist (28).The dis-

46D
Can J Gastroenterol Vol 14 Suppl D November 2000   Usually the H pylori status can be evaluated by means of a rapid urease test and/or histology within the initial endoscopic procedure.In those cases with a negative result, additional serological testing may be recommended with respect to an unknown antibiotic pretreatment or a secondary loss of the bacterium by tumour progression.
The stage of gastric lymphoma is usually described by the Musshoff classification for extranodal lymphoma (29) and its modification of stage EI by Radaszkiewicz et al (26) (Table 2).More recently, another classification system has been proposed -the Lugano classification (30).The prognostically important differentiation of stages EI1, EI2 and EII1 in both classification systems requires the use of endoscopic ultrasound, at least if a nonsurgical approach is favoured.It is the only morphological procedure that is able to visualize the different layers of the gastric wall and the perigastric lymph nodes (Figures 2 and 3).There is some evidence from a small series that success or failure of H pylori eradication therapy can be predicted by endoscopic ultrasound (31).In a prospective study of 77 patients with newly diagnosed gastric lymphoma, endoscopic ultrasound correctly predicted the depth of tumour infiltration (stage EI1 versus EI2) and the lymph node status (stage EI versus EII1) in 78% and 75%, respectively, compared with the gold standard of the pathohistological stage of the resected specimen (32).The main source of error was 'positive' lymph nodes in 21% of patients that were found at histology to be inflammatory reactions and not neoplastic changes.Thus, the diagnostic accuracy of endoscopic ultrasound is good, but not optimal.It may be further improved when endosonographic-guided puncture is available in the future.Also, the value of miniechoendoscopes has to be evaluated in this field.

THERAPY
Options in the treatment of gastric lymphoma include H pylori eradication, surgical resection, radiation, chemotherapy and combinations of these modalities.
The convincing evidence for a pathogenetic role of H pylori infection, as pointed out above, inevitably involved a therapeutic effort.In 1993, Wotherspoon et al (33) reported a complete regression of low grade lymphoma following successful H pylori eradication in five-sixths of cases.Since then, several case reports and some prospective trials have confirmed this observation (33-46) (Table 3).The success rate of this approach is about 80%.Usually, endoscopic and/or histological lymphoma regression occurs within one to three months; however, it may be delayed in individual cases.An interview of an international expert proved that a wait-and-see attitude for six to 18 months is justifiable, provided that there is no sign of disease progression (H Boot, personal communication).The majority of experts define treatment failure as lymphoma being present one year after successful eradication of H pylori.
Undoubtedly, the eradication of H pylori represents a fascinating therapeutic option in low grade lymphoma of stage EI.However, the ultimate value of this principle will need to be further evaluated, because there are still many open questions: Can J Gastroenterol Vol 14 Suppl D November 2000 47D Gastric MALT lymphoma • What types of processes are involved in the lymphoma regression?Apoptosis?Differentiation?
• Which parameters indicate autonomous tumour growth and which criteria can predict the success of eradication therapy (Figure 1)?
• How often does lymphoma relapse?Is recurrence regularly accompanied by H pylori reinfection?
• What is the risk of high malignant transformation?
• What is the significance of persisting monoclonal B cells found in a considerable percentage of patients with histologically, completely regressed lymphoma (46,47)?
These aspects clearly underline that, despite its simplicity and unexpected individual successes (48,49), H pylori eradication therapy in gastric lymphoma needs to be evaluated in clinical trials offering adequate diagnosis and close long term follow-up.Assuming that the disease is cured seems premature.Until very recently, surgery was the standard procedure in gastric lymphoma or was a regular component of combined treatment modalities.The arguments for a surgical approach are as follows: • Surgery provides precise histological classification and staging.
• Surgery is the most frequent treatment modality used so far and is also the treatment that doctors are most familiar with in cases of gastric lymphoma.In a meta-analysis of 80 studies with 3528 patients, a surgical approach was performed in 83% of patients (56).Only 17% of the patients were treated exclusively by radiation and/or chemotherapy.Five-year survival rates were significantly higher in surgically treated patients (60% to 64%) than in those treated with a conservative strategy (34% to 44%).
• Surgery avoids complications, such as perforation or hemorrhage, that may occur in primary radiochemotherapy.This risk has, however, been overestimated for some time now (57-59) and is probably not higher than the postoperative mortality rate.
• Surgery may be the patient's preference.
In recent years, opinion has increasingly swung toward primary radiotherapy or chemotherapy.The arguments for a nonsurgical therapeutic approach follow: • Due to better endoscopic bioptic techniques and the introduction of endoscopic ultrasound, there is no need for surgery for diagnostic reasons.Possible limitations, as pointed out above, have to be kept in mind.
• The possibility of organ preservation contributes to a better quality of life.
• There is evidence from some recent studies for a high efficacy of conservative strategies in gastric lymphoma (57)(58)(59)(60)(61)(62)(63)(64)(65).These have, however, been hardly comparative analyses.The multicentre study by Koch et al (64) left it to each centre's discretion to choose surgical resection or primary conservative treatment in stages EI and EII.With reservations of possible bias, the equivalence of surgical and conservative strategies is nevertheless remarkable.
• Given the risk of extragastric recurrence or dissemination following surgery for high grade lymphoma, systemic chemotherapy is indicated.
• A nonsurgical treatment may be the patient's preference.
The decision for a treatment strategy combining local therapy (surgical resection or radiation) with systemic chemotherapy seems too logical, at least for advanced stages and high grade lymphoma.The majority of studies have chosen such an approach (56).In the German and Austrian prospective multicentre trial, surgery was the initial treatment in stages EI and EII (except for low grade lymphoma of stage EI, which received H pylori eradication therapy).Stratified according to pathohistological stage and grade of malignancy, patients were postoperatively treated by chemotherapy and/or radiotherapy, or entered into follow-up observation.Complete remission was achieved in 96% of patients (28).Based on an intention-to-treat analysis, R0 resection was significantly superior to R1/2 resection with respect to overall survival.

FUTURE THERAPEUTIC ASPECTS
There is no standardized, generally accepted therapy for gastric lymphoma; the search for the best therapeutic approach must continue.The main clinical challenge for the future is to clarify whether surgery remains an option or whether it can be replaced with conservative treatment modalities offering equivalent efficacy and better quality of life.A satisfactory answer to this question can be best provided by a prospective, randomized study.An European multicentre trial was initiated in 1998.

Figure 2 )Figure 3 )
Figure 2) Endoscopic ultrasound showing extension of the lymphoma beyond the organ per continuitatem