You are here
Primary intestinal B-cell lymphoma: A prospective multicentre clinical study of 91 cases
Digestive and Liver Disease, 11, 45, pages 947 - 952
Background & Aims
Clinical presentation, diagnosis and prognosis of different primary intestinal lymphomas have not been well described and compared so far. Our aim was to prospectively analyse a series of consecutive patients presenting different types of B-cell primary intestinal lymphomas.
Adult patients with primary intestinal lymphoma, collected between 1991 and 2000 within the multicenter national study in France were evaluated and followed up prospectively. Clinical features and treatment outcomes were analyzed and compared among different groups of lymphomas.
Among 91 cases of B-cell primary intestinal lymphomas identified, 38 (41%) were diffuse large B cell lymphomas, 34 (37%) mantle cell lymphomas, 12 (13%) follicular lymphomas, 5 (5%) marginal zone MALT-lymphomas, and 2 (3%) Burkitt's lymphomas. A differential diagnosis could be made on the basis of tumour cell morphology and phenotype assessed by immunohistochemistry. Clinical presentation of the different types of lymphomas varied with respect to age, symptoms, circumstances of diagnosis, and stage. Overall survival was the poorest for mantle cell lymphomas while diffuse large B cell lymphomas could be cured if in complete remission after first line treatment.
This study underlines the existence, within the B-cell primary intestinal lymphomas, of several distinct entities with different clinico-pathological features and prognosis, whose identification is important for choosing appropriate therapeutic strategy.
Keywords: Primary intestinal lymphoma, B-cell lymphoma, MALT lymphoma, Follicular lymphoma, DLBCL, Mantle cell lymphoma, Prospective study.
The gastrointestinal tract is the most common site of extra-nodal involvement in non-Hodgkin's lymphomas, the stomach being the most frequently concerned. The primary intestinal lymphomas (PIL), which include lymphomas of the small and large bowel, account for 20 to 30% of all gastrointestinal lymphomas and their incidence is increasing  . PILs differ from gastric lymphomas in clinical features, pathology, treatment and prognosis , , , , and . While a substantial progress has been made in the diagnosis and treatment of gastric lymphoma, allowing to publish recent European guidelines  and , PIL remains a much less characterized entity for which no standardized diagnosis and treatment modalities exist. This is related to the fact that PIL is a rare and heterogeneous disease. Until now, mainly retrospective and/or selective series concerning only some histological subtypes of PIL, or referring to the ancient histological classification, mainly of small size , , , , and , or histological series  and , have been published.
The most recent studies, describing clinico-pathological characteristics of PIL classified according to the new World Health Organization (WHO) lymphoma classification, are mainly retrospective and come from Asia , , , , and .
In the only prospective European study of PIL published until now  , 56 patients with PIL were included but because of selection bias, T-cell lymphomas were over-represented and only 21 patients with B-cell lymphoma were included.
Prospective studies with a sufficient number of patients and accurate histological typing of the B-cell PIL are missing. Therefore, our aim was to describe the clinico-pathological features of different intestinal lymphomas and compare their clinical characteristics and survival. Accordingly, all consecutive patients with diagnosed PIL were included in this prospective multicentre study, and subsequent analysis focused on B-cell lymphomas which are much more common than T-cell lymphomas and represent over 85% of PIL.
2. Patients and Methods
From January 1991 to January 2000, the cases of PIL were collected as a part of a multicenter study on gastrointestinal lymphomas conducted by the French Digestive Lymphoma Study Group (Groupe d’Etude des Lymphomes Digestifs, GELD). Twenty four centers recruited the patients for this study. According to the study chart, data concerning demographic, clinical, biological and histological features, as well as the treatment applied and the clinical outcome, had to be recorded in a prospective manner. Therapeutic guidelines were given and followed for each subtype of lymphoma, but the therapeutic strategy for individual patient was left to the discretion of the investigator and had to be registered. Furthermore, the follow-up data were collected.
The primary character of intestinal lymphoma was defined according to Lewin et al as either predominant intestinal lesion without previous peripheral lymphadenopathy or any lymphoma whose clinical presentation was related to intestinal involvement  . The diagnosis of PIL was based on the WHO classification  . Informed consent was obtained from all the patients and the study was approved by the local Ethics Committee.
Medical history and symptoms were recorded and for each patient the WHO index was evaluated. Initial clinical symptoms were divided into five principal categories: abdominal pain, bloody stools, diarrhoea, abdominal mass, and obstruction.
2.2. Histology and immunohistochemistry
Histology of each tumour, based on endoscopic biopsies and/or surgical material, was reviewed by at least two expert pathologists. Tumours were reclassified according to the WHO classification of tumours of haematopoietic and lymphoid tissues  . Immunohistochemical analysis was performed on paraffin-embedded or fresh frozen biopsies, and included staining with anti-CD20, anti-CD3 and anti-cytokeratin antibodies to confirm the B-cell nature of the lymphoid infiltrate and to visualize lymphoepithelial lesions. Additional staining with anti-CD5, anti-CD10, anti-cyclin D1, anti-bcl-2 and MIB-1 (Ki67) was used to further characterize the tumour  and .
The staging procedure included recording of patients’ complete history, physical examination, routine haematological and biochemical tests (liver function test, full blood count, lactate dehydrogenase, β-2microglobulinemia, and viral serologies), upper and lower endoscopy (with multiple biopsies of the tumour and multilevel systematic biopsies), small bowel and chest X-rays or thoracic CT-Scan, abdominal CT-Scan, Waldeyer's ring examination and bone marrow biopsy. Cerebrospinal fluid was studied in patients with disseminated diffuse large B cell lymphomas and Burkitt's lymphomas. Extension of the disease was defined according to the Ann Arbor staging system, modified by Musshoff  .
2.4. Treatment and response to treatment
Therapeutic guidelines included initial surgical tumour resection only if necessary for emergency or diagnostic purpose. Primary resection was avoided when there was a risk of mortality, or of morbidity delaying chemotherapy, or when the extent of resection was considered excessive with multiple lymphoma localisations. According to the histological type, patients were offered different therapeutic strategies.
Response to treatment was assessed at the end of the initial chemotherapy by a work-up identical to the pre-therapeutic one. Complete remission was defined as complete disappearance of any lesion from all sites initially involved and the absence of new tumours. Further evaluations were planned every year for at least 5 years. Other treatment outcomes were classified, according to the defined criteria as partial remission, stability of the disease, progression of the disease, and death.
2.5. Statistical analysis
Statistical analysis was performed using SAS 9.3. The survival rates were calculated using the Kaplan-Meier method. Overall Survival (OS) was calculated from the date of diagnosis to the date of the final follow-up or death from any cause. Progression Free Survival (PFS) was calculated from the date of diagnosis to the date of disease progression, relapse or to the date of the final follow-up. Comparisons between the groups were tested by the log-rank test.
3.1. Patients and histological diagnosis
From 1991 to 2000, 361 cases of gastrointestinal lymphomas entered the national survey of GELD. Ninety- nine (28%) of them were PILs. From those, 91 (93%) were of B-cell type and 8 (8%) were of T-cell type and thus were not included in the analysis. Data from 91 patients [53 men (58.2%), median age: 56 years, range: 15-78 years] with confirmed primary B-cell intestinal lymphoma were collected. Demographic data of all the patients are presented in Table 1 . According to histological diagnosis and phenotype analysis, five different pathological types of lymphoma could be distinguished: 1) Diffuse large B-cell lymphomas (DLBCL), 38 cases (42%); 2) Mantle cell lymphomas (MCL), 34 cases (37%) (including some cases previously reported: ref 24); 3) Follicular lymphomas, 12 cases (13%); 4) Extra nodal marginal zone lymphomas of mucosa associated lymphoid tissue (MZL- MALT), 5 cases (5%) and 5) Burkitt's lymphomas, 2 cases (3%). DLBCL were made of sheets of large cells with numerous mitoses. MALT lymphomas were composed predominantly of small cells, with scattered large cells. Lymphoepithelial lesions were rare. MCL were composed of monotonous small cells, centrocytic- like, sometimes admixed with scattered epithelioid histiocytes. Follicular lymphomas had follicular pattern and were composed of small cleaved cells (centrocytes) and few large round cells (centroblasts). Burkitt's lymphomas were made of sheets of monomorphic medium-sized cells with numerous mitoses and apoptotic cells.
|Type of lymphoma||No (%)||M/F (Sex ratio)||Age (years) Median (range)|
|DLBCL||38 (42)||20/18 (1.1)||55 (15-73)|
|MCL||34 (37)||22/12 (1.8)||57 (21-78)|
|FL||12 (13)||7/5 (1.4)||56 (40-68)|
|MZL-MALT||5 (5)||3/2 (1.5)||66 (49-74)|
|BL||2 (3)||1/1 (1.0)||31 (17-45)|
|Total||91 (100)||53/38 (1.4)||56 (15-78)|
Abbreviations: DLBCL, diffuse large B-cell lymphoma; MCL, mantle cell lymphoma; FL, follicular lymphoma; MZL-MAL, Tmarginal zone lymphoma-MALT; BL, Burkitt's lymphoma; MMale; F, Female; nnumber of cases
Table 2 shows tumour expression of CD5, CD10, Cyclin D1, bcl-2 and MIB1. A strong and exclusive expression of CD5 and CyclinD1 in MCL was noted. CD10 was expressed in all follicular lymphomas and Burkitt's lymphomas and in some cases of DLBCL. Bcl-2 expression was present in all types of lymphoma with the exception of Burkitt's lymphomas. Proliferative index Ki67 was the highest in Burkitt's lymphomas (100%) and the lowest (5-20%) in MZL-MALT.
|Type of Lymphoma (n)||CD5||CD10||Cyclin D1||Bcl-2||MIB1 (% of positive cells)|
|MZL - MALT (5)||−||−||−||+||5-20|
Abbreviations: DLBCL, diffuse large B-cell lymphoma; MCL, mantle cell lymphoma; FL, follicular lymphoma; MZL-MALT, marginal zone lymphoma-MALT; BL, Burkitt's lymphoma; −negative in all; +, positive in all; −/+, negative or positive; n, number of cases
3.2. Clinical presentation
Abdominal pain was the most frequent initial symptom in all groups (mean 57%, range: 33-60%), either linked to the lymphoma or not, especially in case of follicular lymphoma. Indeed, in this type the diagnosis was often made accidentally during a routine endoscopy for dyspepsia. Intestinal bleeding (18% and 25%) and diarrhoea (20% and 17%) were relatively common in MCL and follicular lymphoma, respectively. Endoscopic diagnosis was the rule in all lymphomas except for DLBCL where diagnosis was often made during surgery (68%) performed for initial complication (obstruction or perforation) ( Table 3 ).
|Type of Lymphoma (n)||Abdominal pain n (%)||Bloody stools n (%)||Abdominal mass n (%)||Obstruction n (%)||Diarrhea n (%)||Surgical diagnosis n (%)|
|DLBCL (38)||18 (47)||1 (3)||2 (5)||15 (40) §||2 (5)||26 (68)|
|MCL (34)||16 (47)||6 (18)||4 (12)||1 (3)||7 (20)||5 (14)|
|FL (12) *||4 (33)||3 (25)||0||2 (17)||2 (17)||3 (25)|
|MZL-MALT (5) *||3 (60)||0||0||0||1 (20)||0|
|BL (2)||1 (50)||1 (50)||0||1 (50)||1 (50)||0|
|Total (91)||52 (57)||11 (12)||6 (6)||19 (21)||13 (14)||34 (37)|
* including 1 asymptomatic patient
§ including 1 perforation
Abbreviations: DLBCL, diffuse large B-cell lymphoma; MCL, mantle cell lymphoma; FL, follicular lymphoma; MZL-MALT, marginal zone lymphoma-MALT; BL, Burkitt's lymphoma; nnumber of cases
DLBCLs were presented mostly as ulcerated tumours while MCL and follicular lymphoma as polypoïd or nodular disseminated lesions (73% and 83% of patients, respectively). Small bowel was the main localisation of DLBCL, follicular lymphoma and MZL-MALT, while colon was the most frequently affected in MCL patients. Within the small bowel, jejunum and ileum were the most frequently involved, followed by ileo-caecal region and duodenum. Multiple intestinal localisations characterised MCL (100%) and follicular lymphoma (67%) patients ( Table 4 ). Moreover, 10 patients presented secondary macroscopic or microscopic gastric localisations: 9 in MCL and 1 in MZL-MALT group.
|Type of lymphoma (n)||Endoscopic aspect||Main sites||Multiple digestive tract segments involved n(%)||Stage I/IIE n (%)|
|Tumor n (%)||Ulcer n (%)||Small bowel n (%)||Colon n (%)|
|DLBLC (38)||33 (86)||1||36 (95)||2 (5)||5 (13)||35 (92)|
|D: n = 4, J + I: n = 23|
|IC: n = 9|
|MCL (34)||9 (26)||0||7 (20)||27(80)||34(100)||10 (29)|
|D: n = 0, J + I: n = 3|
|IC: n = 4|
|FL (12)||1 (8)||1||8 (67)||4 (33)||8 (67)||11 (91)|
|D: n = 3, J + I: n = 4|
|IC: n = 1|
|MZL-MALT (5)||1||1||4 (80)||1 (20)||0||5 (100)|
|D: n = 4, J + I: n = 0|
|IC: n = 1|
|BL (2)||1||1||1 (50)||1 (50)||1 (50)||2 (100)|
|D: n = 0, J + I: n = 1|
|IC: n = 0|
|Total (91)||45 (49)||4 (4)||56 (62)||35 (38)||48 (53)||63 (69)|
|D: n = 11 (12)|
|J + I: n = 31 (34)|
|IC: n = 15 (16)|
Abbreviations: DLBCL, diffuse large B cell lymphoma; MCL, mantle cell lymphoma; FL, follicular lymphoma; MZL-MALT, marginal zone lymphoma–MALT; BL, Burkitt's lymphoma; D, Duodenum; J + Jejunum I + Ileon; IC, Ileo-caecal region; nnumber of cases
Patients with MCL presented more advanced disease at diagnosis than all the other groups, with bone marrow involvement (71% of stage IV)( Table 4 ). LDH and β-2 microglobulin levels were more frequently elevated in MCL (29% and 50% of patients, respectively), than in all other groups of patients taken together (5% and 2%, respectively, p < 0.05).
3.3. Treatment and outcome
According to the histological type of lymphoma, patients were offered different therapeutic strategies according to the guidelines provided for each histological lymphoma subtype.
Among 38 DLBCL patients, 19 were treated by chemotherapy alone: 6 cycles of CHOP regimen (adriamycin 50 mg/m2 i.v. day 1 + vincristine 1.4 mg/m2 i.v. day 1, + cyclophosphamide 750 mg/m2 i.v. day 1 + prednisolone 60 mg/m p.o. day 1 to 5), and 19 with surgery + /- chemotherapy with 4 cycles of CHOP. All 34 MCL patients received 6 courses of CHOP or COP regimen, followed in 13 of them by intensification with autologous bone marrow stem cell transplantation. MCL patients who could not tolerate or who refused intensive therapy were treated by mono-chemotherapy with alkylating agent (cyclophosphamide). In the follicular lymphoma group (12 cases), 4 patients were offered “watch and wait” strategy, two patients had intestinal resection (one for diagnosis and the second one for obstructive tumour), and 6 patients were treated by chemotherapy (6 cycles of COP regimen: cyclophosphamide 750 mg/m2 i.v. day 1 + vincristine 1.4 mg/m2 i.v. day 1 + prednisolone 60 mg. day 1 to 5). All MALT patients were treated by chemotherapy (COP) and additionally, two of them received surgery and two others, radiotherapy. Both patients with Burkitt's lymphoma were treated by chemotherapy (CHOP) combined with intrathecal chemotherapy.
The median follow-up after treatment for the entire cohort was 40 months (range: 1 -175 months). The 5-year OS of the whole population was 77% (95% CI: 63%-86%). The best 5-year OS was achieved in patients with follicular lymphoma (100%), followed by DLBCL [83% (95% CI: 65%-92%)]. Patients with MCL presented the poorest 5-year OS [63% (95% CI: 36%-81%)]. Given a small number of patients in MZL-MALT group, OS in this group has to be interpreted with caution. Out of two patients with Burkitt's lymphoma, one was alive after 5 years while the data concerning the second patient was missing. A non-significant trend to a better 5-year OS was observed in DLBCL patients as compared to MCL patients ( Table 5 , supplementary Fig. S1).
|Type of Lymphoma (n)||Median follow-up months (range)||5-year OS (%)||5-year PFS (%)|
|DLBCL (38)||38 (1 -110)||82||77|
|MCL (34)||39 (4 - 775)||63||39|
|FL (12)||98 (3–177)||100||87|
|MZL- MALT (5)||59 (8–73)||67||67|
|Total (90)||40 (1–775)||77||65|
Abbreviations: DLBCL, diffuse large B-cell lymphoma; MCL, mantle cell lymphoma; FL, follicular lymphoma; MZL-MALT, marginal zone lymphoma-MALT; BL, Burkitt's lymphoma; OS, overall survival; PFS, progression free survival; NE, non evaluated; n, number of cases
The overall 5-year PFS of all patients was 65% (95% CI: 51%-77%), and it was 77% (95% CI: 59%-88%), 39% (95% CI: 15%-63%) and 88% (95% CI: 39%-98%), for patients with DLBCL, MCL and follicular lymphoma, respectively ( Table 5 , supplementary Fig. S2). It can be noted that out of 5 patients with MZL-MALT, 4 were progression-free after 5 years, and that one patient with Burkitt's lymphoma was progression-free after 5 years. A trend towards a better PFS in follicular lymphoma patients, as compared to DLBCL and MCL groups, was observed but the difference was not statistically significant (p = 0.53).
The present study confirms that among all gastro-intestinal lymphomas, stomach is the most frequently involved site, followed by the small and large intestine , , , , , , and . Out of two main types of intestinal lymphomas, B-cell lymphoma is far more frequent (92%) than T-cell lymphoma. This has been already demonstrated by previous studies in which the proportion of B-cell lymphomas among PILs varied from 74 to 100% ( Table 6 ) , , , , , and .
|Author/Year(Ref)||Nakamura et al., 2000(16)||Daum et al., 2003(21)||Li et al., 2008(19)||Wang et al., 2011(20)||Kim et al., 2011(18)||Present study|
|Type of the study #||R/Clin||P/Clin||R/Clin||R/Clin||R/Path||P/Clin|
|B-cell (% of total number of lymphomas)||67 (84)||21 (38)||38 (95)||60 (74)||504 (87)||91 (92)|
|DLBCL n(%)||36 (55) **||18 (86)||17 (45)||35 (58)||386 (77)||38 (42)|
|MCL n(%)||2 (3)||3 (5)||19 (4)||34 (37)|
|FL n(%)||4 (6)||1 (5)||3 (5)||7 (1)||12 (13)|
|MZL-MALT n(%)||15 (22)||2 (10)||20 (53)||15 (25)||61 (12)||5 (5)|
|BL n(%)||7 (10)||4 (6)||31 (6)||2 (3)|
|Age mean (range)||56 (5-89)||57 (21-75)||44 (13-78)||46 (18-75)||56 (15-78)|
|Sex ratio (M/F)||11/10||26/13||42/18||53/38|
|Small intestine n(%)||59||14||352||57 (63%)|
|Colon n(%)||23||106||33 (36%)|
|Multifocal n(%)||46||47 (52%)|
* Only the studies with WHO histological classification of primary intestinal lymphomas are presented
** % of all B-cell lymphoma is given
# R = retrospective, P = prospective, Clin = clinical, Path = pathological only.
Abbreviations: DLBCL, diffuse large B-cell lymphoma; MCL, mantle cell lymphoma; FL, follicular lymphoma; MZL-MALT, marginal zone lymphoma-MALT; BL, Burkitt's lymphoma; Mmale; F, female; OS, overall survival; n, number of cases
Most of the studies published until now reported small and retrospective series, or histological series (with no clinical data) of PIL, and most of them were based on ancient lymphoma classification , , , , , , and . Five studies of PIL diagnosed or re-evaluated according to the new WHO classification could be identified ( Table 6 ) and four of them come from Asia. It is conceivable, however, that Asiatic populations may be different, from epidemiological and clinical point of view, from the European population. Only one European prospective study published until now, showed a serious selection bias and included mostly T-cell lymphomas  . Accordingly, our series appears to be the largest European series of B-cell type of PIL published so far. Moreover the present study was prospective allowing collection of clinical and pathological data in consecutive patients. We chose to focus on B-cell PIL since they are more frequent than T-cell lymphomas, their clinical, biological and prognostic features are not well established, and their clinical management is not standardized.
Among different types of PIL, DLBCL is the most frequent as indicated by several studies where the percentage of DLBCL varied from 45 to 86% , , , , , and . The relatively high proportion of MCL (37%) found in our study does not seem representative for PIL since in most of the published series this proportion was much lower, varying from 3 to 19%. This may reflect a selection bias existing in our institution which is a reference center for gastro-intestinal lymphomas in general and for MCL in particular  . The same explanation probably applies to follicular lymphomas, more common in our series (13%) than in other published studies (1 to 3% of PIL patients)  and . Conversely, the prevalence of MZL-MALT in our series appears lower (5%) than in other reported series (10 to 53%), and we cannot see any clear explanation for this finding. As far as Burkitt's lymphoma is concerned, it is mainly a childhood disease thus its low frequency in our adult population was expected.
The clinical presentation of PIL in our series was dominated by abdominal pain and diarrhoea, followed by bloody stools and abdominal mass, but differed depending on lymphoma localisation and its histological type. Endoscopic diagnosis was the rule in all types of lymphomas, especially those with main colonic involvement (MCL, MZL-MALT), or in patients with follicular lymphoma, where it was often incidental, established during endoscopy performed for dyspeptic symptoms that could not always be attributed to the presence of the tumour. Conversely, in case of DLBCL, which more frequently involved small intestine, surgical diagnosis was frequent and made during surgery for initial complications, mainly ulcerated or obstructive tumour.
The small intestine is the most frequent localization of PIL , , , , and  and it was also confirmed by our study. The only exception was MCL, which was more frequently localized in the colon (76%). MCL and follicular lymphoma were mainly presented in the colon as polypoïd or nodular disseminated lesions (lymphomatous polyposis). The risk of confounding these two types of lymphomas (MCL and follicular lymphoma), showing similar phenotype and both consisting of small or medium -sized B-cells, used to be high in the past but is nowadays much lower since immunohistochemistry allows to differentiate easily these two types of lymphomas (high expression of cyclinD1 in MCL).
More than half of our patients presented multifocal disease, with multiple intestinal sites present at diagnosis, in particular in the group of MCL, as already reported , , , and , and of follicular lymphoma  and , while in most of the patients with DLBCL a single site was involved. Moreover, 9 patients with MCL and 1 patient with MZL-MALT presented secondary gastric localisations. Multiples localizations in PIL have already been underlined by other studies, but their frequency have been usually lower than 50%  and . The higher frequency observed in our study may be attributable to the high number of MCL patients included in our series.
High rate of multifocal localisation in PIL underlines the interest of exploring entire gastrointestinal tract (lower and upper endoscopy, video-capsule endoscopy, etc) mainly for prognostic purposes, in these patients. However, the multifocal character could be difficult to ascertain and may not always have impact on the choice of treatment that will often consist in chemotherapy. Additionally, a PET-Scan may be performed (DLBCL, MCL or follicular lymphoma) in order to evaluate extension of the disease, although it was not used at the time of this study.
Most of the patients (69%) were diagnosed at an early stage of the disease (I–IIE), except for the patients with MCL who in majority (>70%) were diagnosed at an advanced stage, indicating a known higher aggressiveness of this type of lymphoma , , , and .
The optimal treatment strategy in intestinal lymphomas is not established. Patients with multifocal or disseminated disease such as MCL are good candidates for chemotherapy. Accordingly, all our patients with MCL were treated with chemotherapy and 13 of them, with intensification and autologous stem cell transplantation. However, this type of lymphoma remains of poor prognosis (OS = 67%) probably because this intensification is not always possible due to the advanced age of the patients (over 65) at diagnosis. It should be noted, however that chemotherapy regimens have been improved and nowadays, even in older patients with MCL, chemotherapy gives better prognosis than at the time of our study.
Chemotherapy can be also proposed for follicular lymphoma if the tumour mass is important (stage IV or large tumours) while the small and localised tumours could benefit from the “watch and wait” strategy. Immunotherapy (anti-CD20, rituximab), not used at time of this study, should improve prognosis. The same approach could be used with respect to MZL-MALT. For DLBCL and Burkitt's lymphoma, which involve often small intestine with initial obstruction or perforation, despite their well-known chemo-sensitivity, surgery is often mandatory as first line treatment giving in both strategies good prognosis (82%, OS). The result may be further improved by adding immunotherapy.
The best OS was achieved in follicular lymphoma and DLBCL, and the poorest in MCL. This observation is in concordance with previously published reports and with the evidence of high aggressiveness of MCL. Similarly, the best PFS was achieved in follicular lymphoma and the poorest in MCL. It should be emphasized, however, that three out of eight of our cases of follicular lymphoma were localized in the duodenum and it is well known that follicular lymphoma of the duodenum has an extremely good prognosis as compared to that occurring in the rest of the gastro-intestinal tract  . This may have influenced the good overall survival of follicular lymphoma in our series.
This study presents some limitations. The first is a small number of patients, especially in MZL-MALT, follicular lymphoma and Burkitt's lymphoma groups, which makes it difficult to draw firm conclusions in these groups of patients. The second is due to the fact that patients were included from 1991 and 2000 and that considerable progress has been made since that time, especially in the treatment of these lymphomas. In particular, the targeted immunotherapies (rituximab) introduced more recently, have considerably modified the treatment strategies and the outcome of these patients. Accordingly, the treatment modalities evolving from this study may not always be considered current standard.
In summary, this large prospective study describes a large diversity of PIL, which show distinct clinical, pathological and prognostic features. It emphasises the importance of a correct initial diagnosis in which immunochemistry studies are mandatory to assess histological subtype in order to guide the treatment and surveillance of these patients.
Conflict of interest
No conflict of interest has been declared
This work was supported by grants from the Délégation à la Recherche Clinique and Assistance Publique-Hôpitaux de Paris (AP-HP). We wish to thank Ms Lucie Planche for her help in statistical analysis.
-  C. Lepage, A.M. Bouvier, S. Manfredi, et al. Incidence and management of primary malignant small bowel cancers: a well-defined French population study. Am J Gastroenterol. 2006;101:2826-2832 Crossref.
-  B. Dragosics, P. Bauer, T. Radaszkiewicz. Primary gastrointestinal non-Hodgkin's lymphomas. A retrospective clinicopathologic study of 150 cases. Cancer. 1985;55:1060-1073 Crossref.
-  P. Koch, F. del Valle, W.E. Berdel, et al. Primary gastrointestinal non-Hodgkin's lymphoma: I. Anatomic and histologic distribution, clinical features, and survival data of 371 patients registered in the German Multicenter Study GIT NHL 01/92. J Clin Oncol. 2001;19:3861-3873
-  A. Ruskone-Fourmestraux, P. Aegerter, A. Delmer, et al. Primary digestive tract lymphoma: a prospective multicentric study of 91 patients. Groupe d’Etude des Lymphomes Digestifs. Gastroenterology. 1993;105:1662-1671
-  W. Fischbach, B. Dragosics, M.E. Kolve-Goebeler, et al. Primary gastric B-cell lymphoma: results of a prospective multicenter study. The German-Austrian Gastrointestinal Lymphoma Study Group. Gastroenterology. 2000;119:1191-1202 Crossref.
-  S. Nakamura, T. Matsumoto, M. Iida, et al. Primary gastrointestinal lymphoma in Japan: a clinicopathologic analysis of 455 patients with special reference to its time trends. Cancer. 2003;97:2462-2473 Crossref.
-  A. Ruskone-Fourmestraux, W. Fischbach, B.M. Aleman, et al. EGILS consensus report. Gastric extranodal marginal zone B-cell lymphoma of MALT. Gut. 2011;60:747-758 Crossref.
-  E. Zucca. Gastric marginal zone lymphoma of mucosa-associated lymphoid tissue type: ESMO clinical recommendations for diagnosis, treatment and follow-up. Ann Oncol. 2007;18(Suppl 2):ii59-ii60
-  P. Domizio, R.A. Owen, N.A. Shepherd, et al. Primary lymphoma of the small intestine. A clinicopathological study of 119 cases. Am J Surg Pathol. 1993;17:429-442 Crossref.
-  E.M. Ibrahim, A.A. Ezzat, A.N. El-Weshi, et al. Primary intestinal diffuse large B-cell non-Hodgkin's lymphoma: clinical features, management, and prognosis of 66 patients. Ann Oncol. 2001;12:53-58 Crossref.
-  H.F. Gou, J. Zang, M. Jiang, et al. Clinical prognostic analysis of 116 patients with primary intestinal non-Hodgkin lymphoma. Med Oncol. 2012;29:227-234 Crossref.
-  S. Kako, K. Oshima, M. Sato, et al. Clinical outcome in patients with small-intestinal non-Hodgkin lymphoma. Leuk Lymphoma. 2009;50:1618-1624 Crossref.
-  P.L. Zinzani, M. Magagnoli, G. Pagliani, et al. Primary intestinal lymphoma: clinical and therapeutic features of 32 patients. Haematologica. 1997;82:305-308
-  H.D. Foss, H. Stein. Pathology of intestinal lymphomas [In Process Citation]. Recent Results Cancer Res. 2000;156:33-41 Crossref.
-  S. Kohno, K. Ohshima, S. Yoneda, et al. Clinicopathological analysis of 143 primary malignant lymphomas in the small and large intestines based on the new WHO classification. Histopathology. 2003;43:135-143 Crossref.
-  S. Nakamura, T. Matsumoto, M. Takeshita, et al. A clinicopathologic study of primary small intestine lymphoma: prognostic significance of mucosa-associated lymphoid tissue-derived lymphoma. Cancer. 2000;88:286-294 Crossref.
-  S.L. Wang, Z.X. Liao, X.F. Liu, et al. Primary early-stage intestinal and colonic non-Hodgkin's lymphoma: clinical features, management, and outcome of 37 patients. World J Gastroenterol. 2005;11:5905-5909
-  S.J. Kim, C.W. Choi, Y.C. Mun, et al. Multicenter retrospective analysis of 581 patients with primary intestinal non-hodgkin lymphoma from the Consortium for Improving Survival of Lymphoma (CISL). BMC Cancer. 2011;11:321
-  B. Li, Y.K. Shi, X.H. He, et al. Primary non-Hodgkin lymphomas in the small and large intestine: clinicopathological characteristics and management of 40 patients. Int J Hematol. 2008;87:375-381 Crossref.
-  G.B. Wang, G.L. Xu, G.Y. Luo, et al. Primary intestinal non-Hodgkin's lymphoma: a clinicopathologic analysis of 81 patients. World J Gastroenterol. 2011;17:4625-4631
-  S. Daum, R. Ullrich, W. Heise, et al. Intestinal non-Hodgkin's lymphoma: a multicenter prospective clinical study from the German Study Group on Intestinal non-Hodgkin's Lymphoma. J Clin Oncol. 2003;21:2740-2746 Crossref.
-  K.J. Lewin, M. Ranchod, R.F. Dorfman. Lymphomas of the gastrointestinal tract: a study of 117 cases presenting with gastrointestinal disease. Cancer. 1978;42:693-707 Crossref.
-  S.H. Swerdlow, E. Campo, M. Seto, H.K. Muller-Hermelink. SH Swerdlow (Ed.) WHO classification of tumours of haematopoietic and lymphoid tissues (IARC, Lyon, 2008) 229-232
-  A. Ruskone-Fourmestraux, A. Delmer, A. Lavergne, et al. Multiple lymphomatous polyposis of the gastrointestinal tract: prospective clinicopathologic study of 31 cases, Groupe D’etude des Lymphomes Digestifs. Gastroenterology. 1997;112:7-16 Crossref.
-  A. Ruskone-Fourmestraux, J. Audouin. Primary gastrointestinal tract mantle cell lymphoma as multiple lymphomatous polyposis. Best Pract Res Clin Gastroenterol. 2010;24:35-42 Crossref.
-  K. Musshoff. [Clinical staging classification of non-Hodgkin's lymphomas (author's transl)]. Strahlentherapie. 1977;153:218-221
-  K.A. Gurney, R.A. Cartwright, E.A. Gilman. Descriptive epidemiology of gastrointestinal non-Hodgkin's lymphoma in a population-based registry. Br J Cancer. 1999;79:1929-1934 Crossref.
-  R. Otter, R. Bieger, P.M. Kluin, et al. Primary gastrointestinal non-Hodgkin's lymphoma in a population-based registry. Br J Cancer. 1989;60:745-750
-  T. Kodama, K. Ohshima, K. Nomura, et al. Lymphomatous polyposis of the gastrointestinal tract, including mantle cell lymphoma, follicular lymphoma and mucosa-associated lymphoid tissue lymphoma. Histopathology. 2005;47:467-478 Crossref.
-  A. Salar, N. Juanpere, B. Bellosillo, et al. Gastrointestinal involvement in mantle cell lymphoma: a prospective clinic, endoscopic, and pathologic study. Am J Surg Pathol. 2006;30:1274-1280 Crossref.
-  G. Damaj, V. Verkarre, A. Delmer, et al. Primary follicular lymphoma of the gastrointestinal tract: a study of 25 cases and a literature review. Ann Oncol. 2003;14:623-629 Crossref.
-  J. Misdraji, N.L. Harris, R.P. Hasserjian, et al. Primary follicular lymphoma of the gastrointestinal tract. Am J Surg Pathol. 2011;35:1255-1263 Crossref.
-  A.I. Schmatz, B. Streubel, E. Kretschmer-Chott, et al. Primary follicular lymphoma of the duodenum is a distinct mucosal/submucosal variant of follicular lymphoma: a retrospective study of 63 cases. J Clin Oncol.. 2011;29:1445-1451 Crossref.
a Institute of Digestive Diseases (IMAD) and Department of Gastroenterology, Hotel Dieu Hospital, Nantes
b Department of Pathology, Saint-Antoine Hospital, Paris
c Department of Gastroenterology, Saint Antoine Hospital, Paris
d Department of Hematology, Saint-Antoine Hospital, Paris, France
© 2013 Editrice Gastroenterologica Italiana S.r.l., Published by Elsevier B.V.