Treatment of primary gastric lymphoma and gastric mucosa-associated lymphoid tissue lymphoma

Collective Reviews

Treatment of Primary Gastric Lymphoma and GastricMucosa-Associated Lymphoid Tissue Lymphoma

Jeffrey Stephens, MD, and Judy Smith, MD, FACS

Primary gastric lymphoma (PGL) is an uncommonmalignancy that is increasing in incidence and repre-sents approximately 2% to 7% of primary gastricmalignancies.1-5 In the United States there will be anestimated 22,400 cases of gastric malignancy in 1997,which indicates an estimated 448 to 1,568 new cases ofgastric lymphoma will occur.6 Primary gastric lym-phoma is defined as an extranodal lymphoma arising inthe stomach and is the most common site of extranodallymphoma. PGL can spread to regional lymph nodesand become disseminated. Most primary gastric lym-phomas are B cell in origin although occasional casesof T-cell and Hodgkin’s lymphoma are seen. Second-ary gastric lymphoma indicates involvement of thestomach as a part of a diffuse lymphoma arising else-where. In an autopsy series, patients that died of dis-seminated non-Hodgkins lymphoma (NHL) had in-volvement of the gastrointestinal tract in 50% to60% of cases.7 Tertiary gastric lymphoma is recurrentlymphoma involving the stomach after treatment ofa nodal lymphoma and is uncommon.

STAGING AND GRADINGThe Ann Arbor staging system was initially describedfor Hodgkin’s lymphoma, but it is commonly ap-plied to non-Hodgkin’s lymphoma.8 This system isshown in Table 1. Staging of gastric lymphoma basedupon the Ann Arbor system includes stage IE, whichis disease limited to the stomach without nodalspread. Stage IIE1 is tumor in the stomach withspread to adjacent contiguous lymph nodes. StageIIE2 is tumor in the stomach with spread to lymphnodes that are noncontiguous with the primary tumor.

In 1994 Shimodaira and colleagues proposed astaging system for PGL based on the principles of theTNM system.9 This system is shown in Table 2. TheT stage is based on the depth of invasion with inva-sion of adjacent organs included as a staging factor.The N stage is based on the pattern of nodal spreadand takes into consideration the noncontiguous na-ture of nodal spread. The M stage indicates distant ordisseminated disease. Division into stage I throughstage IV shows significant prognostic value. Al-though this system is designed for primary gastriclymphoma, a major deficit is the lack of a variable inthe system to account for tumor grade that has been

Received March 3, 1998; Accepted for publication April 17, 1998.From the Department of Surgical Oncology, Roswell Park Cancer Institute,Buffalo, NY.Correspondence address: Jeff Stephens, MD, 3600 Gaston Avenue, WadleyTower, Suite 958, Dallas, Texas 75246.

Table 1. Ann Arbor Classification for Staging of Hodgkin’s Lymphoma

Stage Area of involvement

I One lymph node regionIE One extralymphatic organ or siteII Two or more lymph node regions on the same side of the diaphragmIIE One extralymphatic organ or site and criteria for stage IIIII Lymph node regions on both sides of the diaphragmIIIE One extralymphatic organ or site and criteria for stage IIIIIIs Spleen and criteria for stage IIIIIISE Spleen and one extralymphatic organ or site and criteria for stage IIIIV One or more extralymphatic organs or sites with or without lymph node involvement, diffuse or disseminated

312© 1998 by the American College of Surgeons ISSN 1072-7515/98/$19.00Published by Elsevier Science Inc. PII S1072-7515(98)00180-X

shown to be an independent predictor of surviv-al.10,11 In a 1997 review of published data from 100series of gastric lymphoma the most common grad-ing system used was based upon the Ann Arbor grad-ing system.12

Results of a workshop on the pathologic andstaging classifications for gastrointestinal lympho-mas was reported in 1994.13 In this report the patho-logic classification as proposed by Isaacson that fol-lowed the principles of the Kiel classification wasrecommended. This is shown in Table 3. In this re-port a uniform pretreatment evaluation was pro-posed as well as a system for grading based upon theAnn Arbor Classification.13

PATIENT PRESENTATIONSymptoms of primary gastric lymphoma that aremost common at presentation are abdominal painand weight loss. Bleeding is less common and pa-

tients rarely present with perforation (Table 4).3,14-18

Stage IE and stage IIE primary gastric lymphomapresent with an equal prevalence. Stage IE disease isseen in 36% to 72% of patients at presentation andstage IIE disease is seen in 28% to 64%.1,3,19-23 Pre-sentation with high grade and low grade disease isalso equal with 34% to 65% presenting as highgrade, 12% to 17% presenting as intermediate gradeand 35% to 65% presenting as low grade lympho-ma.5,10,11,21,24,25

ENDOSCOPIC DIAGNOSISThe endoscopic appearance of gastric lymphoma isvariable and can be flat, polypoid and ulcerated. Pol-ypoid lesions are seen in 25% to 37% of the patients;flat lesions are noted in 18% to 25% of the patients,with the rate of ulceration between 34% to73%.10,11,26 PGL is most often identified in the mid-dle and distal third of the stomach with a significantincidence of multifocal and diffuse disease. Lesionsare found in the distal stomach in 9% to 59%, bodyin 12% to 70%, proximal stomach in 1% to 24%,multiple sites in 6% to 21%, and diffusely in thestomach in 7% to 38%.3,11,14,16,18,26-28

Diagnosis of primary gastric lymphoma by endo-scopic evaluation has markedly improved over the

Table 2A. TNM Staging for Primary GastricLymphoma

Stage Characteristic

T Stage1 Tumor invades the lamina propria or submucosa2 Tumor invades the muscularis propria3 Tumor invades the subserosa4 Tumor penetrates the serosa, no invasion of adjacent

structures5 Tumor invades adjacent structures

N Stage0 Negative nodes1 Positive nodes in perigastric nodes within 3 cm of

the primary tumor2 Positive nodes in perigastric nodes more than 3 cm

from the primary tumor or along the hepatic, leftgastric, splenic, or celiac arteries

3 Positive nodes in the hepatoduodenal, paraaortic,distant abdominal nodes

4 Positive nodes beyond N3M Stage

0 No metastasis1 Distant metastasis positive

Table 2B. TNM Staging for Primary GastricLymphoma

Stage T N M5-year

survival (%)

I T1 N0, N1 M0 100II T1 N2 M0

T2, T3 N0, N1, N2 M0 88.9III T4, T5 Any N M0

Any T N3, N4 M0 52.1IV Any T Any N M1 25

Table 3. Pathologic Classification for GastrointestinalLymphoma from Workshop

B-cell lymphomaLow grade lymphoma of MALTHigh grade lymphoma of MALT—with or without

evidence of low grade componentImmunoproliferative small intestinal diseaseLow grade, mixed of high gradeMantle cell lymphomaBurkitt-like lymphomaOther types of low or high grade lymphoma corresponding

to peripheral lymph node equivalentsT-cell lymphoma

Enteropathy associated T-cell lymphomaOther types not associated with enteropathy

MALT, mucosa-associated lymphoid tissue

Table 4. Primary Gastric Lymphoma Symptoms atPresentation

SymptomsIncidence

(%)

Abdominal pain 71–85Weight loss 11–68Nausea/emesis 14–28Bleeding 7–23Perforation 0–6

313Vol. 187, No. 3, September 1998 Stephens and Smith Treatment of Gastric Lymphoma

past decade.3 In recent series successful endoscopicdiagnosis of lymphoma was made in 92% to 100% ofthe patients.10,28 A second endoscopy is still neededto make the diagnosis in a few patients. A large reviewof 3,157 patients with primary gastric lymphomareported data on the sensitivity of endoscopic diag-nosis for 848 patients.12 A diagnosis of cancer wasobtained in 79% with a diagnosis of lymphoma ob-tained in only 54%. This indicates that in these pa-tients 25% of the patients were diagnosed with can-cer not believed to be lymphoma. They also statedthat the success of endoscopic evaluation and biopsyto diagnose gastric lymphoma did improve markedlyin recent decades. In a review of patients treated from1980 until 1990 the correct preoperative diagnosiswas made in only 57% of patients but an improve-ment over time was noted. A diagnosis rate of 44%was seen between 1980 and 1984 compared with64% from 1985–1990.3 Other series report a correctdiagnosis by endoscopy in 38% to 92%.14,15,17,18,26

Techniques that have lead to a higher rate of successinclude the use of larger biopsy forceps, multiple bi-opsies, and the use of brushings. If an ulcer is present,the biopsy should be at the edge of the ulcer crater atmultiple sites. The role of endoscopic ultrasound indiagnosis and staging of primary gastric lymphoma isunknown at this time although it may allow moreaccurate pretreatment staging.

TREATMENT OF EARLY STAGE PRIMARYGASTRIC LYMPHOMATreatment of primary gastric lymphoma is highlyvariable with no consensus established on the mosteffective therapy. To date there has been only onerandomized clinical trial published comparing treat-ments. Treatment decisions must therefore be basedon retrospective studies that address the treatment ofthis uncommon malignancy. These retrospectivestudies are difficult to compare and mostly representnonrandomized data with significant stage and gradevariation. Selection bias is evident in some series;patients with more advanced disease and more ag-gressive histology receive more aggressive therapy.

Overall 5-year survival of stage IE and IIE pri-mary gastric lymphoma is from 52% to93%.5,11,15,21-23,25,28 The 5-year survival for stage IEranges from 46% to 91%, for IIE from 44% to 82%,and is not statistically different in most se-ries.1,3,10,11,14 The effect of grade is a more significantfactor in relation to survival. Survival for low gradePGL is 75% to 95% and for high grade 46% to 56%

at 5 years.3,10,11 Combining stage and grade may im-prove the prognostic ability. In a review by Cogliattiand associates the survival was 95% for those patientswith low grade stage IE tumors, 82% for low gradestage IIE tumors, 74% for high grade stage IE tu-mors, and 39% for high grade Stage IIE tumors.11

Treatments which have been used for primarygastric lymphoma include surgery, radiation, andchemotherapy as single modality therapy. Multimo-dality therapy has included surgery with adjuvantradiation, surgery with adjuvant chemotherapy, sur-gery with both radiation and chemotherapy, or radi-ation with chemotherapy (without surgery). A recentarticle which included 3,528 patients from the liter-ature reviewed the variety of treatments that havebeen used.12 The majority of the patients receivedsurgery either alone or in combination and showedan improved survival over patients treated withoutsurgery. This may represent a selection bias withmore advanced cases and unresectable cases beingtreated with chemotherapy and radiation.

Surgical therapy is felt by some to be the primaryform of treatment and is thought to be necessary foraccurate staging, grading, and to protect the patientfrom the risk of perforation and bleeding duringtreatment with radiation and chemotherapy. Recentseries has shown the incidence of perforation andbleeding with medical treatment is extremely lowand ranges from 0% to 7%.20,24,30 This risk is equiv-alent to surgical mortality (0% to 9%).17,18,21,23,25

The incidence of late morbidity such as malnutritionand dumping are more common in patients treatedwith gastrectomy.25 Recent series have demonstratedorgan preservation with nonsurgical treatment withequivalent longterm survival compared with regi-mens which include surgical resection.20,25

PROSPECTIVE RANDOMIZEDTREATMENT RESULTSIn 1996 the results of a prospective randomized trialwere reported. A total of 75 patients were random-ized between 3 treatments consisting of surgeryalone, surgery with adjuvant chemotherapy, or pre-operative radiation followed by surgery and adjuvantchemotherapy.23 Survival at 3 years was 51% withsurgery alone, 79% with surgery and chemotherapy,and 100% with radiation followed by surgery andchemotherapy. All of these differences were statisti-cally significant. The patients were equal with respectto stage and grade and no significant difference wasseen in the incidence of morbidity and mortality. It

314 Stephens and Smith Treatment of Gastric Lymphoma J Am Coll Surg

should be noted that this prospective randomizedstudy contains no nonsurgical treatment arm forcomparison.

RETROSPECTIVE TREATMENT DATA

Primary surgical therapyIn some retrospective series, treatment with surgeryalone has been shown to be effective with no benefitseen with the use of adjuvant therapy.Treatment withsurgery alone is most successful in patients with earlystage disease. In a review of 34 patients with stage IEand IIE1 gastric lymphoma, patients were treatedwith complete surgical resection and lymph nodedissection. There were 15 patients treated with sur-gical resection alone compared to 19 patients treatedwith surgery and adjuvant therapy (chemotherapy, 8;radiation, 6; radiation and chemotherapy, 5).14

There was no difference in survival suggesting noadditional benefit with the addition of adjuvanttreatment. A separate review of 84 patients found nosurvival benefit in patients treated with postoperativeradiation following curative or palliative surgerywhen compared with surgery alone.27 Schwarz andcoworkers reviewed 56 patients with stage IE and IIEPGL and found that complete surgical resection wasa significant factor related to outcomes but the addi-tion of radiation therapy had no effect on survival.3

Successful treatment with surgery alone in stageIE and IIE disease was also shown by Zinzani andcolleagues, who demonstrated a 10-year survival of93% in patients treated with surgery alone when thedisease was limited to the serosa, the lymph nodeswere negative, and the margins of resection wereclear.21 This group noted equal survival for patientswho were found to have either invasion through theserosa, positive lymph nodes, positive margins, or acombination of these when they were treated withsurgery and adjuvant radiotherapy. This may suggesta role for the use of adjuvant radiation in patientswith more advanced regional disease. A recent pro-spective study of 50 cases of stage IE and IIE PGLwas reported by Sano and associates.28 In this seriesall patients had surgical resection with patients foundto have positive nodes receiving chemotherapy. The5-year survival was 86% with no postoperative mor-tality. They concluded that primary surgery with ad-juvant chemotherapy for selected cases is appropriatetreatment for PGL. There was no patient treatmentarm without surgery for comparison. Cogliatti andcoworkers reported a series of 145 stage IE and IIEpatients that were all treated with surgery with 65

patients receiving adjuvant therapy (22 radiation, 10chemotherapy, 33 both).11 There was no differencein survival with relation to the type of therapy givensuggesting no benefit with the addition of adjuvanttherapy over surgery alone. Each of these series showno benefit to adjuvant therapy when compared withsurgery alone, yet none of these series in early stagePGL contain a nonsurgical treatment arm forcomparison.

Primary nonsurgical therapyNonsurgical therapy has also been shown to be effec-tive in the treatment of early PGL. Surgical resectionmay offer no survival benefit and in fact may beunnecessary is some patients. Burgers and colleaguesreported the treatment results of 38 patients withstage IE and IIE PGL.29 There were 21 stage IE pa-tients; of these 7 received radiation only and 14 re-ceived surgery and radiation. There were 17 stage IIEpatients; of these 7 were treated with radiation andchemotherapy and 10 were treated with surgery andradiation. Disease-free survival at 5 years was 85% inboth surgical and nonsurgical groups. In additionthey noted a complete response in 6 of 7 stage IEpatients with radiation alone. They also noted nosignificant difference in survival of stage IIE patientswhen treated with surgery and radiation comparedwith nonsurgical treatment with radiation and che-motherapy. They concluded that surgery was notnecessary in the treatment of early stage primary gas-tric lymphoma.

Evidence of the effectiveness of chemotherapyalone in the treatment of gastric lymphoma wasshown by Tanaka and associates. They demonstratedthat there was no residual tumor in 5 of 5 surgicalspecimens after preoperative chemotherapy.30 A pro-spective nonrandomized trial containing 52 Stage IEand IIE patients by Aviles and colleagues compared28 patients treated with chemotherapy alone with 24patients treated with surgical resection followed bychemotherapy.They showed no difference in survivalbetween these 2 groups and concluded that surgery isnot necessary.31

Preservation of the stomach and good survivalwas achieved by Maor and coworkers. Thirty-fourstage IE and IIE patients were treated with radiationand chemotherapy, without surgery, with a 5-yearsurvival of 73%. There was no reported incidence ofbleeding or perforation.20 Two patients received asalvage gastrectomy which led to stomach conserva-tion in 24 of the 26 surviving patients. Followup


ranged from 12 to 40 months with no recurrences inthe patients who had organ preservation.

Finally, a retrospective review by Brinkner andassociates compared 106 patients with stage IE andstage IIE1 PGL treated with 7 different regimens in-cluding both surgical and nonsurgical therapy.25 Thetreatment included 18 with surgery alone, 16 withradiation alone, 12 with chemotherapy alone, 18with surgery and radiation, 18 with surgery and che-motherapy, 13 with surgery and both radiation andchemotherapy, and 8 with radiation and chemother-apy alone. Overall there was no difference in survivalbetween these treatment groups. Also no differencein survival was seen when comparing patients treatedwith and without surgery, concluding that surgicaltreatment was not necessary. This study shows effec-tive therapy with no significant difference in survivalusing all types of therapy both alone and in combi-nation for early stage PGL.

Multimodality therapyOthers promote the use of multimodality therapy inthe treatment of primary gastric lymphoma. Vali-centi and colleagues reviewed 77 stage IE and IIEpatients who were treated with a variety of therapiesand showed a significantly improved survival whenpatients were treated with multimodality therapycompared with those treated with single modalitytherapy (65% versus 24%).32 In contrast they re-ported no difference in survival in patients with tu-mors , 5 cm when treated with radiation alonecompared with treatment with surgery and radiation.In 1993 Pasini and associates reported a series of 53stage IE and IIE patients, all of whom had surgery.22

Postoperative chemotherapy was given in high riskpatients. A significant improvement in disease-freesurvival was seen in those who received the chemo-therapy (92% versus 60%) despite the fact that pa-tients treated with chemotherapy had more advanceddisease.20 The incidence of relapse was higher inthose treated initially with surgery alone (32% versus3%) but 4 of the 6 patients who relapsed after surgeryalone were rescued with chemotherapy. Because ofthe effectiveness of chemotherapy rescue there wasno difference seen in overall survival.

In summary, the treatment of primary gastriclymphoma is highly variable. In early disease the useof single modality therapy with surgery, chemother-apy, and radiation have all been shown to be poten-tially curative in retrospective reviews. There is noconsensus on the treatment of intermediate stage dis-ease, ie stage IIE or stage I high grade, but it is

thought to be best treated with chemotherapy withsurgery and radiation being variable.

MUCOSA ASSOCIATED LYMPHOMATISSUE (MALT)Isaacson introduced the concept that low grade gas-tric lymphoma resembled histologic features of mu-cosa associated lymphoid tissue (MALT) rather thanthat of nodal lymphoid tissue.33 Mucosa associatedlymphoid tissue is found normally in the small bowelin the form of Peyers patches, Waldeyers ring, appen-dix and bronchus. Normally there is no lymphoidtissue associated with the stomach and therefore par-adoxical that this is the most common site of extra-nodal NHL. Before the formation of lymphoma inthe stomach it must first acquire lymphoid tissue. Inthe stomach the appearance of MALT is believed tobe in response to infection with the bacteria Helico-bacter pylori.7,34

GASTRIC MALT LYMPHOMAThis accumulation of lymphoid tissue in the stom-ach is believed to be the precursor of the formation ofgastric lymphoma in most cases. Most cases of pri-mary gastric lymphoma are of B-cell origin andclosely resemble MALT rather than nodal lymphoidtissue. The connection of gastric MALT lymphomaand infection with H. pylori has been well estab-lished.35,36 It has been shown that H. pylori is presentin the stomach of patients with gastric lymphoma ina majority of cases.16,35 An epidemiological associa-tion between the incidence of H. pylori infection andthe incidence of gastric lymphoma was shown byDoglioni and coworkers in 1992.37 They found thatgeographic regions with a higher incidence of gastriclymphoma also had a higher incidence of infectionwith H. pylori.

More convincing evidence of this association wasshown by Parsonnet and colleagues in a case con-trolled cohort study which demonstrated that pa-tients with gastric lymphoma were more likely tohave evidence of infection with H. pylori than thematched controls (RR 6.3).38 This report also statedthat infection with H. pylori preceded the onset ofgastric lymphoma.

Hussel and associates in 1992 demonstrated theresponse of MALT lymphoma cells to H. pylori andfound the response to be strain specific.39 Theyfound that both T cells and neoplastic B cells showedan increase in proliferation and an increase in Inter-leukin 2(IL-2) production in response to exposure to


H. pylori antigen. They showed that removal of thenonneoplastic T cells resulted in a reduced prolifera-tion and IL-2 production. This illustrated that theproliferation of the monoclonal malignant B cellswas dependent upon the presence of nonmalignant Tcells and their production of IL-2.

Current concepts of the link between the infec-tion by H. pylori and the development of gastric lym-phoma are described in a recent report by Isaacson.40

He suggests that after the accumulation of lymphoidcells in the stomach in response to H. pylori infection,genetic abnormalities in B cells, possibly trisomny 3,lead to the development of a monoclonal prolifera-tion and low grade lymphoma. This proliferation ofmonocolonal B cells is thought to remain dependentupon the antigenic stimulation by H. pylori and theproduction of IL-2 by nonneoplastic T cells. Thiswas previously felt to represent what was termedpseudolymphoma but is now considered true lym-phoma with the findings of uniform gene rearrange-ment and the presence of monoclonal antibody pro-duction.40 The B-cell proliferation is stimulated bynonmalignant T cells which are producing IL-2 inresponse to the bacterial antigen. Further genetic re-arrangements lead to the development of B-cell pro-liferation or low grade lymphoma which is indepen-dent H. pylori antigen. Finally low grade lymphomaprogresses into high grade lymphoma. Evidence forthe final transformation is suggested with the com-mon occurrence of combined high grade and lowgrade lymphoma in a surgical specimen.

HISTOLOGIC DEFINITION—LOW GRADEMALT LYMPHOMAHistologic grading for MALT lymphoma was de-scribed by Isaacson.34,40 Low grade MALT lym-phoma is histologically similar to Peyers patches. Re-active nonneoplastic follicles are present and thelymphoma cells invade around and between the re-active follicles corresponding to the area of the Peyerspatch marginal zone. They invade individual gastricglands forming the characteristic lymphoepitheliallesions. These lympoepithelial lesions can be seen innonneoplastic MALT in association with H. pyloriinfection but do not have any other features of thelymphoma. Cellular appearance is variable but is of-ten described as small to medium sized cells withabundant cytoplasm and irregular nuclei that havethe appearance of centrocytes. Low grade lesions canbe multifocal and can be associated with plasma celldifferentiation and scattered transformed blasts.

HISTOLOGIC DEFINITION—HIGHGRADE MALT LYMPHOMAHigh grade MALT lymphoma cannot be reliably dif-ferentiated from other high grade B cell lymphomas.Foci of high grade lesions are often seen in low gradetumors and vice versa.35,41 The presence of confluentclusters of sheets of transformed cells outside of thecolonized follicles represents the high grade MALTlymphoma. The reactive follicles and lymphoepithe-lial lesions of low grade lymphoma are less commonand may not be seen.

MALT LYMPHOMA PRESENTATIONSignificant differences are seen in the presentation oflow grade MALT lymphoma when compared withhigh grade MALT lymphoma.10,24 Low grade MALTlymphoma is seen more frequently in males and inyounger patients. It is associated with a significantlylower incidence of weight loss, epigastric pain, palpa-ble nodes, hepatomegaly, palpable abdominal mass,and elevated serum LDH. This is reflected in thepresentation of low grade MALT lymphoma with alower incidence of bulky disease and advanced stage.Low grade MALT lymphoma is also associated with alower incidence of polypoid tumor mass and ulcer atendoscopy. Low grade lymphoma is also found tohave smaller tumors with less transgastric invasion,less invasion of adjacent organs, and fewer positivelymph nodes.10,24

Clinical behaviorSignificant differences are seen in the clinical behav-ior of MALT lymphoma when compared to NHL ofnodal origin. The prognosis of MALT lymphoma isgenerally more favorable than that of the equivalentNHL of nodal origin. Low grade MALT lymphomais usually localized and slow to spread to lymph nodesand other sites. Survival of patients with low gradeMALT lymphoma is better than that of patients withhigh grade MALT lymphoma.41 Some reports sug-gest that high grade MALT lymphoma has a morefavorable clinical course than its equivalent highgrade nodal NHL19, but other reports show no suchdifference.42

Treatment of low grade MALT lymphomaStrong evidence exists that supports the idea that thefirst treatment option for low grade gastric lym-phoma should be the eradication of H. pylori. Thishas been shown to be highly successful with regres-sion of the low grade MALT lymphoma in mostcases. Initial evidence came from Wotherspoon and


coworkers in 1993.41 In this study, 5 of 6 patients withlow grade MALT gastric lymphoma had complete re-gression of the lymphoma following successful treat-ment of the H. pylori with antibiotics. A second study byBayerdorffer and associates confirmed these results with33 patients.43 They demonstrated regression of theMALT lymphoma both with routine histologic stainingand with polymerase chain reaction analysis. Afterconfirmed eradication of H. pylori there was com-plete regression of lymphoma in 23 patients (70%)and in 13 of 16 patients tested with polymerase chainreaction. Partial response was seen in 4 (12%) and nochange was seen in 6 (18%). Of the 6 patients thatdid not respond, 1 was treated with chemotherapyand 5 underwent surgery, 4 of whom were found tohave high grade B-cell lymphoma and 1 was found tohave high grade T-cell lymphoma. At 1 year followupthere were no recurrences.

Three recent series report success in treatment oflow grade gastric B cell lymphoma. In the first ofthese reports 50 patients with stage IE low gradegastric MALT lymphoma were treated for H. pyloriwith a cure rate for the bacteria of 100%.44 Followupwas 24 months; 40 patients (80%) had a completeresponse, 6 had no response, and 4 had a partialresponse. In those patients who did not have a re-sponse it was found that 5 of 6 had a high gradecomponent to the tumor. There were 5 patients witha relapse (12.5%). Of 31 patients with complete his-tologic response, 22 were found to have monoclonalB cells using polymerase chain reaction analysis. This

raises the question of whether these patients are curedof lymphoma or only in remission. Further followupis needed. A second study also showed a completeresponse of 80%.45 In the third study of 45 patientseradication of H. pylori was successful in 44 patients(97%).46 Median followup was 22 months, and 30patients had a complete response (67%) and 2 of the30 patients relapsed (7%). The median time to com-plete response in those who did respond was 5months. Obviously these patients need close long-term followup. Given the indolent nature of lowgrade MALT lymphoma and the success of regressionwith eradication of H. pylori, treatment with antibi-otic regimens to eliminate H. pylori should be thefirst line of therapy in the treatment of low gradeMALT lymphoma. Patients that do not respond andthose with high grade disease should have standardtherapy of the treating institution. In a recent publi-cation the National Cancer Care Network publishedsuggested guidelines for the treatment of PrimaryGastric Lymphoma which includes the use of antibi-otic eradication of H. pylori as first line therapy.47

Failures and patients with no evidence of H. pyloriinfection are treated with radiation, with chemother-apy as salvage therapy for radiation failure. The use ofsurgery for the treatment of gastric malt lymphoma isnot included in the guidelines.

CONCLUSIONSPrimary gastric lymphoma is an uncommon diseasewith an increasing incidence. There is a need for pro-spective randomized trials to evaluate the differenttreatments of this malignancy. Evidence for a strongassociation with gastric infection by Helicobacter py-lori has been shown. Evidence also exists for eradica-tion of H. pylori as the first line therapy in the treat-ment of low grade MALT lymphoma. After review ofthe literature we have created a treatment algorithmthat outlines what we believe is a reasonable approachto the treatment of PGL. Treatment of low gradelymphoma is shown in Figure 1.The initial therapy isbased upon type of PGL. Low grade MALT lym-phoma is treated initially with H. pylori eradication.Complete regression following proven eradication isfollowed with serial endoscopies. For patients withstage IE and IIE1 nonMALT lymphoma and lowgrade MALT lymphoma not responding to treat-ment of H. pylori, surgical resection is indicated.Surgical resection has the advantage of allowingaccurate staging, grading, and cure with surgeryalone. This allows chemotherapy to be given only to

Figure 1. Treatment of low grade primary gastric lymphoma. LN,lymph nodes; MALT, mucosa-associated lymphoid tissue.


those patients with high grade lymphoma and avoidschemotherapy in those patients with favorable histol-ogy. Chemotherapy after surgical resection is indi-cated in those patients with positive lymph nodes,residual disease, and high grade lymphoma found inthe pathologic specimen. We feel that all high gradelymphomas, both MALT and nonMALT in origin,should be treated initially with chemotherapy. This isshown in Figure 2. This will effectively treat the lym-phoma and in patients with a complete response willallow for gastric preservation.

References1. Ben-Yosef R, Hoppe RT. Treatment of early stage gastric lym-

phoma. J Surg Onc 1994;57:78–86.2. Kitamura K, Yamaguchi T, Okamoto K, et al. Early gastric lym-

phoma: A clinicopathological study of ten patients, Literaturereview, and comparison with early gastric adenocarcinoma. Can-cer 1996;77(5):850–857.

3. Schwarz RJ, Conners JM, Schmidt N. Diagnosis and manage-ment of stage IE and stage IIE gastric lymphomas. Am J Surg1993;165:561–565.

4. Frazee R, Roberts J. Gastric lymphoma treatment: Medical versusSurgical. Surg Clin N Am 1992;72(2):423–431.

5. Rigacci L, Bellesi G, Alterini R, et al. Combined surgery andchemotherapy in primary gastric non-Hodgkins lymphoma: Aretrospective study of sixty-six patients. Leuk Lymphoma1994;14:483–489.

6. Parker SL, Tong T, Bolden S, et al. Cancer statistics CA 1997;47(1):5–28.

7. Haber DA, Mayer RJ. Primary gastrointestinal lymphoma. SemSurg Onc 1988;15(2):154–169.

8. Pezner RD, Molina A: Non-Hodgkin’s lymphoma in CancerManagement: A Multidisciplinary Approach ed. Pazdur RD, CoiaLR, Hoskins WJ, Wagman LD, PRR Hintington NY, 1996;226–286.

9. Shimodaira M, Tsukamoto Y, Niwa Y, et al. A proposed stagingsystem for primary gastric lymphoma. Cancer 1994;73:2709–2715.

10 Taal BG, Boot H, VanHeerde P, et al. Primary non-Hodgkin lym-phoma of the stomach: Endoscopic pattern and prognosis in lowversus high grade malignancy in relation to the MALT concept.Gut 1996;39:556–561.

11. Cogliatti SB, Schmid U, Schumacher URS, et al. Primary B-cellgastric lymphoma: A clinicopathological study of 145 patients.Gastroen 1991;101:1159–1170.

12. Brands F, Monig SP, Raab M. Treatment and prognosis of gastriclymphoma. Eur J Surg 1997;163:803–813.

13. Rohatiner A, d’Amore F, Coiffer B, et al. Report on a workshopconvened to discuss the pathological and staging classifications ofgastrointestinal tract lymphoma. Ann Onc 1994;5:397–400.

14. Bartlett DL, Karpeh MS, Flippa DA, et al. Long term follow-upafter curative surgery for early gastric lymphoma., Ann Surg1996;223(1):53–62.

15. Pytel J, Sigon R, Bidoli E, et al. Primary gastric non-Hodgkinslymphoma—Does surgery still play any role. Eur J Surg Onc1994;20:525–536.

16. Talanonti MS, Dawes LG, Joehl RJ, Nahrwold DL. Gastrointes-tinal Lymphoma: A case for primary surgical resection. Arch Surg1990;125:972–977.

17. Jones RE, Willis S, Innes DJ, Wanabo HJ. Primary gastric lym-phoma: Problems in staging and management. Amer J Surg1988;155:118–123.

18. Sutherland AG, Kennedy M, Anderson DN, et al. Gastric lym-phoma in Grampian Region: Presentation, treatment and out-come. J R. Coll Surg Edinb 1996;41:143–147.

19. Weingrad DN, Decosse JJ, Sherlock P, et al. Primary gastrointes-tinal lymphoma: A 30 year review. Cancer 1982;49:1258–1265.

20. Maor MH, Velasquez WS, Fuller LM, Silvermintz KB. Stomachconservation in stages IE and IIE gastric non-Hodgkins lym-phoma. J Clin Oncol 1990;8:266–271.

21. Zinzani PL, Frezza G, Bendandi M, et al. Primary gastric lym-phoma: A clinical and therapeutic evaluation of 82 patients. LeukLymphoma 1995;19:461–466.

22. Pasini F, Ambrosetti A, Sabbioni R, et al. Postoperative chemo-therapy increases the disease-free survival rate in primary gastriclymphomas stage IE and IIE. Eur J Cancer 1994;30A(1):33–36.

23. Shchepotin IB, Stephen D, Evans RT, et al. Primary non-Hodgkins lymphoma of the stomach: Three radical modalities oftreatment in 75 patients. Ann Surg Onc 1996;3(3):277–284.

24. Montalban C, Castrillo JM, Abraira V, et al. Gastric B-cellmucosa-associated-lymphoid-tissue (MALT). Clinicopathologi-cal study and evaluation of the prognostic factors in 143 patients.Ann Onc 1995;6(4):798–799.

25. Brinkner H, D’Amore F, et al. A retrospective analysis of treatmentoutcome in 106 cases of localized gastric non-Hodgkin lympho-mas, Leuk Lymphoma 1995;18:281–288.

26. Swaroop S, Mohandas KM, Swaroop VD, et al. Comparativeendoscopic study of primary gastric lymphoma vs. gastric carci-noma. J Surg Onc 1994;56:94–97.

27. Rosen CB, Van Heerden JA, Martin JK, et al. Is an aggressivesurgical approach to the patient with gastric lymphoma war-ranted. Ann Surg 1987;205(6):634–640.

28. Sano T, Sasako M, Kinoshita T, et al. Total gastrectomy for pri-mary gastric lymphoma at stages IE and IIE: A prospective studyof fifty cases. Surgery 1997;121:501–505.

29. Burgers JM, Taal BG, Van Heerde P, et al. Treatment results ofprimary stage I and II non-Hodgkins lymphoma of the stomach.Radiother Oncol 1988;11:319–326.

30. Tanaka Y, Takao T, Watanabe H, et al. Early stage gastric lym-phoma: is operation essential? W J Surg 1994;18:896–899.

31. Aviles A, Diza-Maqueo JC, Del La Torre A, et al. Is surgery nec-essary in the treatment of primary gastric non-Hodgkins lym-phoma? Leuk Lymphoma 1991;5:365–369.

32. Valicenti RK, Wasserman TH, Kuclk NA. Analysis of prognosticfactors in localized gastric lymphoma: The importance of bulkdiseases. J Rad One Biol Phys 1993;27:591–598.

33. Isaacson PG, Spencer J. Malignant lymphoma of mucosa-associated lymphoid tissue. Histopathology 1987;11:445–462.

Figure 2. Treatment of high grade primary gastric lymphoma.MALT, mucosa-associated lymphoid tissue.


34. Isaacson PG. Gastrointestinal lymphoma. Hum Pathol1994;25(10):1020–1029.

35. Isaacson PG. Gastric lymphoma and Helicobacter pylori. N EngJ Med 1994;330(18):1310–1311.

36. Wotherspoon AC, Ortiz-Hildalgo C, Falzon MF, Isaacson PG:Helicobacter Pylori associated gastritis and primary B-cell gastriclymphoma. Lancet 1991;338:1175–1176.

37. Doglioni C, Wothersopoon AC, Moschini A, et al. High inci-dence of primary gastric lymphoma in northeastern Italy. Lancet1992;339:834–385.

38. Parsonnet J, Hansen S, Rodriguez L, et al. Helicobacter pyloriinfection and gastric lymphoma. N Engl J Med 1994;330(18):1267–1271.

39. Hussell T, Isaacson PG, Crabtree JE, Spencer J. The response ofcells from low-grade B-cell gastric lymphomas of mucosa-associated lymphoid tissue to Helicobacter pylori. Lancet1993;342:571–574.

40. Isaacson PG. Recent developments in our understanding of gastriclymphomas. Am J Pathol 1996;20(supp 1):S1–S6.

41. Wotherspoon AC, Doglioni C, Diss TC, et al. Regression of pri-mary low-grade B-cell gastric lymphoma of mucosa-associated

lymphoid type after eradication of Helicobacter pylori. Lancet1993;342:575–577.

42. Morton JE, Leyland MJ, Vaughan Hudaon G, et al. Primary gas-trointestinal Non-Hodgkins’ Lymphoma: A review of 175 Britishnational lymphoma investigation cases. Br J Ca 1993;67:776–782.

43. Bayerdorffer E, Neubauer A, Rudolph B, et al. Regression of pri-mary gastric lymphoma of mucosa-associated lymphoid tissuetype after cure of Helicobacter pylori infection. Lancet 1995; 345:1591–1594.

44. Neubauer A, Thiede C, Morgner A, et al. Cure of Helicobacterpylori infection and duration of remission of low-grade gastricmucosa associated lymphoid tissue lymphoma. J Nat Cancer Inst1997;89(18):1350–1355.

45. Thiede C, Morgner A, Alpen B, et al. What role does Helicobacterpylori eradication play in gastric MALT and gastric MALT lym-phoma. Gastroenterology 1997;113:s61–s64.

46. Pinotti G, Zucca E, Roggero E, et al. Clinical features, treatmentand outcome in a series of 93 patients with low grade gastricMALT lymphoma. Leuk Lymphoma 1997;26: 527–537.

47. NCCN proceedings. Oncology 1997;11(11a):304–315.


Documents

Treatment of primary gastric lymphoma and gastric mucosa-associated lymphoid tissue lymphoma