• endosonography
• mucosa associated lymphoid tissue
• gastric lymphoma

In this issue of Gut, Püspök and colleagues¹ highlight the follow up in patients with primary gastric lymphoma after various stomach conserving therapies: Helicobacter pylori eradication, chemotherapy, radiotherapy, or combinations thereof [see page 691]. Endoscopic ultrasound (EUS) findings were compared with histology and showed that a complete histological remission occurred earlier and more often than an EUS defined remission. Were these results to be expected when treatment response is followed? What are the implications of this study for daily practice?

In the recently formalised WHO classification,² gastric lymphoma is considered as the “disease entity” of marginal zone lymphoma (mucosa associated lymphoid tissue (MALT)-type) with a characteristic morphological spectrum, immunological marker pattern, and discriminative chromosomal aberrations. In the past, these lymphomas were often described as low grade (gastric) MALT lymphomas. High grade (MALT) lymphomas are not a separate entity in the WHO classification but are classified as diffuse large B cell lymphomas (DLBCL) with or without areas of marginal zone lymphoma of the MALT type.

Clinical presentation and endoscopic findings in primary gastric lymphoma may mimic benign and other malignant gastric disorders. The diagnosis of lymphoma can be established on endoscopic biopsies in more than 90% of patients but exact subtyping of the lymphoma and the “horizontal” extension of the disease on a limited number of small endoscopic biopsies may be more difficult.^3–5 Therefore, a second gastroscopy for “intragastric staging” with an extensive biopsy protocol is strongly advocated using a treatment endoscope (for example, Olympus GIF-1T or 2T 100-160) and a large calibre biopsy forceps over the routine endoscopes, as used by others (fig 1). Gastric marginal zone lymphoma of the MALT-type is most often a multifocal disease. Significant foci of malignant large B cells (high grade disease) can be observed in 10–15% of cases^4,5 (de Jong and Boot unpublished). The need for a standardised protocol taking preferably 20–30 biopsies of both abnormal and seemingly uninvolved mucosa is therefore obvious in order to assess horizontal extension and multifocality, both initially and during follow up. Moreover, this will preclude overlooking minor components of transformed/high grade disease that have a negative prognostic impact, both in conventionally treated patients (radiotherapy, chemotherapy) and in patients treated with H pylori eradication.^6,7

Endoscopic ultrasound is beyond dispute the best technique for locoregional staging—that is, assessment of the “vertical” extension of gastric lymphomas—although it is still not employed in all patients.⁸ In our experience, upstaging was shown in 18% (4/22) of patients with low grade MALT lymphomas and in 36% (4/11) of patients with DLBCL with EUS based staging compared with computed tomography based staging. EUS however is not a sensitive technique to diagnose multifocality and horizontal extension of gastric MALT lymphoma.

Apart from the diagnostic value of EUS for locoregional staging, this technique also gives prognostic information on the probability that a histological remission may be obtained after H pylori eradication. This response in gastric MALT-type lymphoma (low grade disease) is strongly related to the extent of involvement of the gastric wall, as shown by endoscopic ultrasound.⁵ If the lesions are confined to the mucosa and submucosa without lymph node involvement, in ±80% of MALT-lymphomas (low-grade) a histological remission may be obtained.⁵ After H pylori eradication, persistence of monoclonal bands has been described in otherwise unsuspicious basal lymphoid aggregates, despite the fact that histological remission has been obtained,⁹ indicating the deep mucosa and submucosa (second and third layer at conventional endoscopic ultrasonography) as regions of special interest. Non-responsiveness to H pylori eradication has been attributed to the presence of DLBCL or high grade disease in the deeper layers of the gastric wall.¹⁰ Recently, the presence of the t(11;18) translocation was shown to be highly predictive of responsiveness to H pylori eradication. Histological complete remission was observed in 68% of 67 translocation negative cases, while in 44 translocation positive cases only two patients showed a transient histological remission.¹¹

During follow up after H pylori eradication in (low grade) MALT lymphoma, endoscopic abnormalities such as ulcers, erosions, and abnormal mucosal folds show marked improvement long before histological remission is documented. In these cases a prolonged follow up of 12–18 months may be necessary, at which time endoscopy is often (near) normal.

In most patients treated with radiotherapy, a radiation dose of 35–40 Gy in daily fractions of 1.8–2 Gy is applied. Assessment of the response with endoscopic biopsies should not be done too early as the response to radiotherapy may also occur slowly. We perform endoscopy two months after the end of radiotherapy. During prolonged follow up after radiation therapy, retraction of the lesser curve may occur due to fibrosis, especially if the primary lesion was located at the angulus or lesser curvature of the antrum.

Patients with DLBCL with or without MALT features can be treated with CHOP-like chemotherapy (cyclophosphamide, doxorubicin, vincristine, and prednisone). A response to chemotherapy in DLBCL may occur very rapidly.

The role of EUS during follow up is not yet determined and the timely study of Püspök et al addresses this topic. They described increased echogenicity during follow up EUS. We performed EUS during follow up of 26 patients and found a normal pattern in five, increased echogenicity in 14 (related to histologically confirmed fibrosis and scarring in 12 after radiotherapy), and increased hypoechoic thickening of the second and third layers in seven patients. In six of these patients recurrent or residual lymphoma after treatment was demonstrated histologically.

The implications of the study of Püspök et al are not easy to translate into clinical practice. In primary gastric lymphoma the response to H pylori eradication may occur more slowly compared with chemotherapy. The value of EUS for follow up in gastric lymphoma seems limited and endoscopy with multiple biopsies remains the investigation of choice. If the first “staging” EUS shows thickened gastric wall layers, it seems prudent to use EUS during follow up after six months, especially when H pylori eradication is used as the primary treatment. A component of DLBCL may be present in the deeper layers of the gastric wall, necessitating treatment adjustment.¹⁰ Moreover, EUS remains useful when endoscopic abnormalities are detected during follow up, especially thickened mucosal folds. Conventional histology remains the “gold standard” for documentation of a complete remission, as the significance of persistent monoclonality in patients with a histological remission remains unclear.⁹

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Depth of biopsy reached by a routine endoscopic biopsy forceps (left) and a large calibre forceps (right). The latter samples well into the muscularis mucosae and upper submucosa.