ORIGINAL ARTICLE

["F] FDG PET in Gastric Non-Hodgkin's Lymphoma Miriam Rodriguez, HBkan Ahlstrom, Anders Sundin, Suzanne Rehn, Christer Sundstrom, Hans Hagberg and Bengt Glimelius

From the Departments of Diagnostic Radiology (M. Rodriguez, H. Ahlstrom), Oncology (S. Rehn, H. Hagberg, B. Glimelius), Pathology (C. Sundstrorn) and the PET Centre (H. Ahlstrom, A. Sundin), University of Uppsala, Akademiska Sjukhuset, S-751 85 Uppsala, Sweden

Correspondence to: Dr Miriam Rodriguez, Department of Diagnostic Radiology, Akademiska Sjukhuset, S-741-85 Uppsala, Sweden. Tel: +46 18 66 30 00. Fax: +46 18 55 72 79

Acta Oncologica Vol. 36, No. 6, pp. 577-584, 1997

The possibility of using ['*F] FDG PET for assessment of tumor extension in primary gastric non-Hodgkin's lymphoma (NHL) was studied in 8 patients (6 high-grade and 2 low-grade, one of the MALT type) and in a control group of 7 patients (5 patients with NHL without clinical signs of gastric involvement, 1 patient with NHL and benign gastric ulcer and 1 patient with adenocarcinoma of the stomach). All patients with gastric NHL and the two with benign gastric ulcer and adenocarcinoma, respectively, underwent endoscopy including multiple biopsies for histopathological diagnosis. All patients with high-grade and one of the two with low-grade NHL and the patient with adenocarcinoma displayed high gastric uptake of [ISF] FDG corresponding to the pathological findings at endoscopy and/or CT. No pathological tracer uptake was seen in the patient with low-grade gastric NHL of the MALT type. In 6/8 patients with gastric NHL, ["F] FDG PET demonstrated larger tumor extension in the stomach than was found at endoscopy, and there was high tracer uptake in the stomach in two patients who were evaluated as normal on CT. ["F] FDG PET correctly excluded gastric NHL in the patient with a benign gastric ulcer and in the patients with NHL without clinical signs of gastric involvement. Although the experience is as yet limited, ["F] FDG PET affords a novel possibility for evaluation of gastric NHL and would seem valuable as a complement to endoscopy and CT in selected patients, where the technique can yield additional information decisive for the choice of therapy.

Received 22 January 1997 Accepted 21 May 1997

The stomach is the most common extranodal site of pri- mary non-Hodgkin's lymphoma (NHL) (1). Although the incidence of gastric NHL has increased in many industrial- ized countries over recent decades, its occurrence is still rare (2, 3). The causes of the disease and of the increase are unknown, but occupational exposure to organic sol- vents and pesticides, as well as Helicobacter pylori (Hp) infection, has been reported as a pathogenetic factor (4).

The diagnosis and staging of gastric NHL is difficult (5) , and the management of the disease has been controversial (6). Various treatment strategies have been proposed, but there is n o consensus regarding the optimal therapeutic approach (6-8). The occurrence and the degree of extra- gastric spread and gastric extension of the disease, together with the malignancy grade, are probably decisive factors in the choice of the therapy. Low-grade NHL of the mucosa- associated lymphoid tissue (MALT) type is, if limited to the stomach, increasingly treated primarily with specific therapy for an assumed underlying H p infection (7, 9, lo), whereas for other types of gastric NHL, surgery, radio- therapy and/or chemotherapy are chosen.

At our hospital, patients with high-grade NHL limited t o the gastric wall are submitted to subtotal gastric resec- tion if it seems likely, during the preoperative workup, that radical surgery will be achieved by this procedure. In all, other patients, primary chemotherapy followed by radio- therapy, or radiotherapy alone, is chosen. In view of the necesssity of these different treatment strategies, there is a need for reliable diagnostic methods which will accurately disclose the tumor extension in the gastric wall.

Available conventional techniques such as preoperative roentgenography of the upper gastrointestinal tract (UGI), endoscopy and computed tomography (CT), have not proven satisfactory for proper diagnosis and staging of gastric N H L (5 , 11). The value of ["F] FDG PET in the detection and malignancy grading of NHL has been re- ported in several studies (12-15). To the best of our knowledge there are no reports, however, on the use of ["F] FDG PET in gastric NHL. The aim of this study was to evaluate the technique in the diagnostic work-up of patients with gastric N H L in comparison with endoscopy and CT.

0 Scandinavian University Press 1997. ISSN 0284- 186X Acta Oncologica

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.

578 M. Rodriguez et al. Acta Oncologica 36 (1997)

Table 1 Characteristics of the patients, examinations performed and diagnosis

Patient Sex/age Diagnosis Tumor extent Malignancy Histology Stage Endoscopy CT MRI PETHs PETHs grade Kiel suv suv

stomach extra-

NHL gastric

1. 2. 3. 4.

5.

6. I .

8.

9.

10.

11.

12.

13. 14. 15.

F/48 M/8 1 F/65 MI58

MI68

MI51 MI75

MI78

MI64

F/80

MI77

F/49

MI72 F/61 MI68

NHL NHL NHL NHL

NHL

NHL NHL

NHL

Carcinoma

NHL and benign ulcer NHL

NHL

NHL NHL NHL

Stomach Stomach Stomach Stomach and LN neck Stomach and LN retro- peritoneal and retrocrural Stomach Stomach

Stomach and LN retrocrural Stomach and liver metastases LN neck

LN thorax and abdomen LN abdomen, neck and thorax. Spleen, liver and bone marrow infiltration LN abdomen LN abdomen LN abdomen

High-grade CB High-grade NUD High-grade CB High-grade NUD

High-grade NUD

High-grade NUD Low-grade IC

Low-grade MALT

- -

High-grade CB

High-grade CB

High-grade CB

High-grade CB High-grade ALCL High-grade CB

IEA + S + G G nd IEA + S + G G + C nd IEA + 8 m m + C nd IEA ++ 8 m m + C nd

IEA ++ S + G G + C nd

IEA + S + G G + C nd IIIEA + S + G G + C nd

IIEA ++ 8 m m nd

- + 8 m m + C nd

IIA + 8 m + C nd

IVB nd nd + IVA nd 8 m m + C nd

IIB nd 8 m m + C nd IIB nd nd + IIA nd nd +

6.2 -

38.6 -

16.5 -

16.6 -

38.7 -

17.1 12.8 -

NPTU NPTU

18.0 -

NPTU -

-

NPTU 13.8

NPTU 10.8

NPTU 9.6 NPTU 10.1 NPTU 16.5

Patients 1-8 gastric non-Hodgkin’s lymphoma (NHL); patients 9- 14 control group; NUD = unclassifiable; LN = pathologically enlarged lymph nodes; CB = diffuse centroblastic lymphoma; IC = immunocytic lymphoma; ALCL = anaplastic large cell lymphoma; MALT =

mucosa-associated lymphoid tissue; + =examination performed; nd = not done; S = spiral CT technique; 8 mm = 8 mm slice CT; GG = 10 tablets of Gastroluft and 1 mg glucagon i.v. before CT; C = i.v. contrast; NPTU = no pathological tracer uptake.

MATERIAL AND METHODS

Patients

The study comprised 15 patients in two groups (Table 1). The first group (patients 1 to 8) consisted of 8 patients aged 48 to 81 (mean 66) years with untreated primary gastric NHL. Extra-gastric spread was known to have occurred in two of the patients. The histopathological diagnosis in this group was based on multiple biopsies a t endoscopy and showed 6 high-grade and 2 low-grade (one of the MALT type) N H L according to the G e l classifica- tion (16). A second group of 7 patients (patients 9 to 15) aged 49 to 80 (mean 67) years served as a control group and included 1 patient with a high-grade NHL in the lymph nodes in the neck and mediastinum and with sus- pected involvement of the stomach, which later proved to be a benign gastric ulcer, 1 patient with gastric adenocar- cinoma and 5 patients with high-grade NHL in the abdo-

men without clinical signs of gastric involvement. The patients with gastric lymphoma were referred to the De- partment of Oncology between April 1992 and March 1995. Tumor staging was performed according to the Ann Arbor system (17). The research protocol was approved by the local ethical committee.

Endoscopy

With the exception of the five control patients with NHL who had no clinical signs of gastric involvement, all pa- tients underwent endoscopy. In 3 patients (2 high-grade and 1 low-grade NHL), the endoscopy was repeated, since a conclusive diagnosis could not be made on the basis of the specimens obtained a t the first examination.

Computed tomography

A total of 12 CT examinations of the upper abdomen was

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.

Acta Oncologica 36 (1997) PET in gastric N H L 579

performed, on a Siemens Somatom Plus (Siemens, Ger- many) (Table 1). In 7 cases an 8 mm slice thickness and increment was used. In 5 cases a spiral CT technique was used, and in order to optimize the examinations, 10 tablets of GastroluftR were given orally and 1 mg (1 Iu) GlucagonR was injected intravenously before scanning. All CT examinations with the exception of two were per- formed with i.v. administration of contrast medium.

Magnetic resonance imaging

Three patients from the control group with high-grade NHL without clinical signs of gastric involvement under- went magnetic resonance tomography (MRI) instead of CT of the upper abdomen. MR imaging was performed with an 0.5 T superconductive unit (Magnetom; Siemens, Erlangen, Germany). Transverse images were obtained with TZweighted and T1-weighted sequences before and after i.v. administration of gadolinium diethylenetriamine- pentaacetic acid (Gd-DTPA; Schering AG, Berlin).

Positron emission tomography

All PET examinations were carried out at the Uppsala University PET Centre with a whole body PET camera, GE 4096 (General Electric medical systems), which pro- duces 15 simultaneous contiguous axial slices with a thick- ness of 6.5 mm and in a plane resolution of 5-6 mm (18). The interval between [18F] FDG PET and CT or MRI was never longer than 12 days. The patients with high-grade NHL without clinical signs of gastric involvement under- went [18F] FDG PET of the upper abdomen with at least two-thirds of the stomach included in the examination. All patients fasted for at least 4 h before examination. After a 1 0-min transmission scan, approximately 400 MBq [ "F] FDG (19) was administered i.v. and a 50 min dynamic scanning sequence was started. This sequence consisted of 14 time frames with successively increasing acquisition times from 1 to 10 min. During the examination, 13 plasma samples were drawn from a peripheral vein of the foot 'arterialized' by warming (20), and analyzed for their "F concentration.

Image reconstruction and analysis

The image reconstruction produced a set of dynamic im- ages, each representing a quantitative estimate of the ra- dioactivity concentration. A 128 x 128 matrix and a 6 mm Hanning filter were used. The images were corrected for attenuation and scattered radiation (21) and data obtained between 30 and 50 min after injection were summed to produce an average image. This average image was then recalculated to provide images of standardized uptake values (SUV), whereby the radioactivity concentration in each pixel was divided by the ratio between the totally administered radioactivity and body weight. In the SUV images, regions of interest (ROI) representing the tumor

area with the highest radioactivity were drawn according to a standardized procedure (15), where an isocontour was positioned halfway between the highest tumor radioactiv- ity and the immediately surrounding tissues. From each ROI obtained in this manner (designated mean ROI), the mean tumor SUV and transport rate were obtained. Also, within each tumor, an additional ROI was created, com- prising the four contiguous pixels with the highest activity and designated 'Hot spot' (Hs).

Evaluation

Endoscopy with biopsies was performed prior to CT and PET investigations. In order to describe the localization of the pathological findings at the various examinations, a schematic diagram of the stomach was drawn (Fig. 1). The stomach was divided into three regions (A, B, C). In each region there is a minor curvature (mi) and a greater curvature (g) area represented by half of the ventral and dorsal gastric wall belonging to each curvature. CT, MRI and PET images were evaluated together for morphologi- cal correlation with the pathological findings. At CT, a thickened gastric wall was judged as tumor infiltration. At PET, areas of high tracer uptake, as compared with sur- rounding normal tissues, were considered as indicating the presence of malignancy.

RESULTS

Endoscopy

At endoscopy, all patients with gastric NHL displayed macroscopic changes, 2 showed visible tumors, and 6 patients were found to have different lesions (infiltrative,

Fig. 1. Diagram of the stomach showing the different regions used for the description of the pathological findings at the various examinations. A: distal portion of the esophagus, cardia, fundus; B: corpus; C: antrum, canalis, bulbus; g: great curvature (shaded area); and mi: minor curvature (white area) represented by the curvatures with half of the ventral and half of the dorsal gastric wall adjacent to each curvature.

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.

580 M, Rodriguez et al. Acta Oncologica 36 (1997)

Fig. 2. Axial CT and PET images of the upper abdomen at corresponding levels in a patient with high-grade gastric non-Hodgkin's lymphoma. (a) CT image (obtained after i.v. injection of contrast medium) shows thickened gastric wall (arrows) corresponding to the corpus-antrum level. K = kidney, L = liver, S = spleen. (b) '*FDG PET SUV image visualizes high tracer uptake corrksponding to the thickened gastric wall (arrows) shown at CT. K =kidney. Red and yellow colors correspond to high tracer uptake.

polypoid or ulcerative), suspected as being malignant. In all 8 patients histopathology was conclusive for gastric NHL (6 high-grade and 2 low-grade). In the two patients in the control group who underwent endoscopy, an ade- nocarcinoma and a benign ulcer, respectively, were confi- rmed at histopathological examination (Table 1).

Computed tomography and magnetic resonance imaging

Of the 8 patients with gastric NHL, CT showed a thick- sened gastric wall in 6 patients (one of whom had a suspected tumor and 2 had pathologically enlarged lymph nodes, in one case both retroperitoneal and retro- crural and in the other cage retrocrural), and a normal gastric wall in the remaining 2 patients. In the control group CT visualized the tumor and liver metastases in the patient with a gastric adenocarcinoma, and a thick- ened gastric wall in the patient with a benign gastric ulcer. In the remaining control patients, CT (or MRI) of the stomach was normal (Table 1).

Positron emission tomography

On PET, high tracer uptake was found in 718 patients with gastric NHL (Fig. 2) and in the patient with an adenocarcinoma (Table 2). In the gastric NHL group, PET demonstrated a larger tumor extent than was found by endoscopy in 6 patients, and showed pathological gastric tracer accumulation in 2 patients in whom CT of the upper abdomen was normal. In one case, where the involved fmdus area was not investigated at PET, the tumor extension was smaller than was demonstrated at

endoscopy and CT. No focal gastric FDG accumulation was found in the patient with low-grade NHL of the MALT type, in the patient with the benign gastric ulcer (Fig. 3) or in those with high-grade NHL without clini- cal signs of gastric involvement. Two patients in the gas- tric NHL group (patients 5 and 8, Table 1) and the patient with adenocarcinoma showed extra-gastric spread on CT. ["F] FDG PET showed high tracer uptake corre- sponding to the extra-gas-tric spread in one of the 2 patients with gastric NHL and the patient with adeno- carcinoma (enlarged retroperitoneal lymph nodes and liver metastases, respectively) (Fig. 4). No pathological tracer uptake was found, however, at a site correspond- ing to the pathologically enlarged retrocrural lymph nodes in the patient with low-grade gastric NHL of the MALT type.

PET meafurements

SUVs obtained at the hot-spot measurements are pre- sented in Table 1. In the patients with gastric high-grade NHL, S U V ranged between 6.2 and 38.7 (mean 25.5), in the patient with low-grade gastric NHL (patient No. 7) 12.8, and in the patient with gastric adenocarcinoma the S U V was 18. No pathological gastric tracer uptake was observed in the patient with low-grade gastric NHL of the MALT type.

DISCUSSION

Our study has shown that [18F] FDG PET is able to demonstrate both the presence and the extension of gas-

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.

Acta Oncologica 36 (1997) PET in gastric N H L 581

Table 2 Regions of pathological changes at endoscopy, CT and PET

Patient Endoscopy CT PET

1. 2. 3. 4. 5. 6. 7. 8. 9.

10.

A, B Cmi Cmi Cmi A A Ami Ami Ami Ami

A, Bg Bmi, C N (**I A, B, C Ami, Bmi + M A, B N A, B g + M Arni + M Ag

Bg ("1 Bmi, C C A, B, C Ami, B m i f M A, B h i , Bmi N Ami + M N

Patients 1-8, gastric NHL; patient 9, gastric adenocarcinoma; patient 10, extra-gastric NHL and benign gastric ulcer; A = distal portion of the esophagus, cardia, fundus; B=corpus; C = antrum, canalis, bulbus; g = greater curvature; mi = minor curva- ture; M =metastasis; N = no pathological changes; (*) = fundus area was not investigated; (**) = difficult to evaluate because of insufficient distension of the stomach.

tric NHL. The extension of the disease and the malig- nancy grade are decisive for the choice of therapy. Gas- trointestinal lymphoma originates in the submucosa and the ability of the different diagnostic methods to detect the disease is dependent on the orientation of the tumor infiltration relative to the gastric wall and the extent of this infiltration. In two studies the currently used meth- ods were not considered sufficiently accurate to establish the diagnosis of the tumor and its eventual extension (5, 11), although both those reports date from more than a decade ago. Before 1970 the histological diagnosis was usually established at major surgery, i.e. a partial or total gastrectomy. The introduction of endoscopic biopsy and CT has decreased the need for surgery to confirm the diagnosis and stage of disease in the patient. The histopathological diagnosis after endoscopic biopsies has shown varying degrees of accuracy. In a retrospective study of 79 patients with gastric lymphoma of Ann Ar- bor stages IE and IIE, Maor et al. (22) concluded that histopathological diagnosis could be obtained by endo- scopic biopsies, avoiding diagnostic laparotomy. Another evaluation of the clinical findings over a 20-year period, and in particular the endoscopic diagnosis, in a series of 119 consecutive patients with primary NHL, stages IE and IIE, showed endoscopic biopsies to be diagnostic in 92% of the cases (23). Poorer detections rates have been reported, however (5, 24).

Endoscopy is thus a useful method for the detection of pathological lesions in the stomach, although it may be difficult to differentiate lymphoma from carcinoma (5) or a benign from a malignant ulcer by this means. In the

present study endoscopy with histopathological examina- tion was conclusive in all cases (in 2/10 cases after re- peated endoscopy).

Although limited by the number of patients, our re- sults suggest that ['*F] FDG PET is a more accurate method than endoscopy and CT for evaluation of the extension of NHL in the gastric wall. For instance, in the group of patients with gastric NHL, high tracer up- take was observed in 2 patients in whom CT scans of the upper abdomen were normal, and in 6/8 patients PET showed a larger tumor extension than was found at endoscopy. Although we do not have histopathological proof that the larger area of tracer uptake in these 6 patients did in fact represent lymphoma, this seems rea- sonable. Theoretically, this tracer accumulation may also reflect an inflammatory reaction in the area circumscrib- ing the lymphoma, but this seems less likely, since it was not seen in the other patients or in the control patient with a benign ulcer. The exact correlation of the uptake in the patient with adenocarcinoma also supports this interpretation. However, we have to consider the fact that the endoscopic examinations were not standardized. We did not repeat the endoscopy, mainly for ethical reasons. The thickened gastric wall on CT, used for the diagnosis of NHL in this study, is an unspecific sign which has been described in cases of benign ulcer, gastric adenocarcinoma, hypertrophic gastritis and other intra- mural gastric tumors (1 1).

Since it was impossible to identify the normal gastric- wall, the tracer uptake could not be measured. For this reason a comparison of pathologic and normal gastric wall uptakes values was not possible. Besides the meth- ods evaluated in this study, two other techniques have been used for staging of gastric NHL, namely endoscopic ultrasonography (EUS) and 67Ga-scintigraphy. In a study comprising 10 patients with gastric NHL, the EUS tech- nique enabled the disclosure of intramural infiltration and perigastric lymph node involvement and was more accurate than endoscopy, barium meal or CT in detect- ing and staging NHL ef the stomach (25). Distant lymph node abnormalities were better visualized by CT, how- ever.

From a scintigraphic study (26), in which 67Ga images of 24 cases of gastrointestinal NHL were reviewed, it was concluded that 67Ga imaging is a useful technique for detecting and monitoring the therapeutic response, but that the absence of 67Ga uptake after therapy is not necessarily a sign of a good therapeutic effect. These results are supported by those of other studies (26-29), but there are reports indicating that 67Ga uptake in the stomach is not specific for NHL and is just as likely to occur in adenocarcinoma, gastritis, and even in a normal stomach (30, 31).

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.

582 M. Rodriguez et at. Acta Oncologica 36 (1997)

Fig. 3. Axial CT and PET images of the upper abdomen at corresponding levels in a patient with high-grade NHL in the neck and benign gastric ulcer. (a) CT image (obtained after oral and i.v. administration of contrast medium) shows thickened gastric wall (arrows) corresponding to the major curvature. L = liver, S = spleen. (b) "FDG PET SUV image. There is no high tracer uptake at the stomach (arrows) compared with surrounding normal tissues. L = liver, S = spleen.

Two previous reports (14, 15) indicate that ["F] FDG PET can predict the malignancy grade in NHL, with the highest uptake values occurring in the group of high-grade NHL. Since the present study included only 6 patients with high-grade NHL and 2 with low-grade NHL, it is not possible to draw any further conclusions in this respect, although the SUVs were within the range that has been reported previously (15). From a clinical point of view, the lack of tracer uptake in the stomach and in the pathologi- cally enlarged retrocrural lymph nodes in the patient with low-grade NHL of the MALT type in the present study is interesting. This was the only patient in whom a lymphoma manifestation demonstrated by any technique did not show pathological tracer uptake with ["F] FDG. The fact that Hp infection may have etiological relevance in the gastric MALT lymphomas suggests a different kind of therapy for patients with low-grade MALT-type gastric NHL, based on antibiotics and antacids (9). The high uptake values obtained in the case of adenocarcinoma reflect the limitations of this technique in differentiating NHL from other gastric malignancies.

We have recommended surgery alone as the primary treatment only for those patients with high-grade and low-grade NHL without extra-gastric spread and in whom radical surgery is likely to be achieved by a subtotal gastric resection; where a total gastrectomy is apparently necessary

because of the location and/or extension of the lymphoma, we have not recommended surgery, since if intensive chemotherapy is judged necessary after this operation, these patients tolerate it poorly. Instead, such patients have been recommended primary chemotherapy with three or four courses of CHOP (cyclophosphamide, doxorubicin, vincristine and prednisone), followed by radiotherapy.

Since in some patients an evaluation of the tumor extension in the gastric wall is fundamental for the choice of therapy, it is important to find reliable diag- nostic methods. Endoscopy is always needed to verify the diagnosis, while CT and MRI provide the necessary morphological information. ["F] FDG PET is an exclu- sive and expensive method, but in selected patients with primary gastric NHL, in whom the extension is difficult to assess, we believe that this technique may give addi- tional information that can have an impact on the choice of therapy. The issue therefore is not whether ["F] FDG PET can replace endoscopy or CT, but rather whether it can yield important additional clinical information.

ACKNOWLEDGEMENTS

This work was supported by grants from the Swedish Cancer Society and from the University Hospital of Uppsala.

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.

Acta Oncologica 36 (1997) PET in gastric NHL 583

Fig. 4. Axial CT and PET images of the upper abdomen at corresponding levels in a patient with high-grade gastric NHL with extra-gastric spread. (a) CT image (obtained after i.v. injection of contrast medium) shows thickened gastric wall (arrows) corresponding to the minor curvature at the corpus level. There is retroperitoneal and retrocrural extra-gastric spread (arrow heads). L = liver, S = spleen. (b) 18FDG PET mass influx image shows high tracer uptake corresponding to the thickened gastric wall and the retrocrural and retroperitoneal extra-gastric spread seen on CT. Red and yellow colors correspond to high tracer uptake.

REFERENCES 1. Aisenberg AC. Coherent view of non-Hodgkin lymphoma. J

Clin Oncol 1995; 13: 2656-75. 2. Pinotti G, Novario R, Berrino F, et al. Primary gastric

non-Hodgkin’s lymphoma in a population-based registry. Haematologica 1992; 77: 405-12.

3. Doglioni C, Wotherspoon AC, Moschini A, de Boni M, Isaacson PG. High incidence of primary gastric lymphoma in northeastern Italy. Lancet 1992; i: 834-5.

4. Fagioli F, Rigolin GM, Cuneo A, et al. Primary gastric lymphoma: distribution and clinical relevance of different epidemiological factors. Haematologica 1994; 79: 21 3-7.

5. Mittal B, Wasserman TH, Griffith RC. Non-Hodgkin lymphoma of the stomach. American J Gastroenterology

6. Rossi A, Rohatiner AZS, Lister TA. Primary gastrointesti- nal non-Hodgkin lymphoma: Still an unresolved question? Annals Oncol 1993; 4: 802-3.

7. Thieblemont C, Berger F, Coiffier B. Mucosa-associated lymphoid tissue lymphomas. Current Opinion in Oncol 1995; 7: 415-20.

8. Di Marco A, Aitini E, Rizzotti A, Grandinetti A, Cam- prostrini F, Smerieri F. Primary non-Hodgkin lymphomas of the gastrointestinal tract: analysis of 41 cases. Tumori 1990; 76: 379-84.

9. Wotherspoon AC, Doglioni C, Diss TC, et al. Regression of primary low-grade B-cell gastric lymphoma of mucosa-asso- ciated lymphoid tissue type after eradication of Helicobacter pylori. Lancet 1993; 342: 575-7.

10. Stoke M, Eidt S. Healing gastric MALT lymphomas by eradicating H pylori? Lancet 1993; 342: 568.

11. Buy JN, Moss AA. Computed tomography of gastric lymphoma. AJR 1982; 138: 859-65.

12. Okada J, Yoshikawa K, Itami M, et al. Positron emission tomography using fluorine 18-fluorodeoxyglucose in malig- nant lymphoma: a comparison with proliferative activity. J Nucl Med 1992; 33: 325-9.

13. Okada J, Yoshikawa K, Imazeki K, et al. The use of FDG- PET in the detection and management of malignant lymphoma: correlation of uptake with prognosis. J Nucl

1983; 78: 780-7.

Med 1991; 32: 686-91. 14. Leskinen-Kallio S, Ruotsalainen U, Ndgren K, Teras M,

Joensuu H. Uptake of carbon-I I-methionine and fluo- rodeoxyglucose in non-Hodgkin’s lymphoma: a PET study. J Nucl Med 1991; 32: 1211-8.

15. Rodriguez M, Rehn S , Ahlstrom H, Sundstrom C, Glimelius B. Predicting malignancy grade with PET in non- Hodgkin’s lymphoma. J Nucl Med 1995; 36: 1790-6.

16. Carbone PP, Kaplan HS, Musshoff K., Smithers DW, Tubi- ana M. Report of the committee in Hodgkin’s disease stag- ing classification. Cancer Res 1971; 31: 1860-1.

17. Lennert K. Malignant lymphomas other than Hodgkin’s dis- ease. In: Handbuch der speziellen plathologischen Anatomie und Histologie. Berlin/Heidelberg/New York: Springer-Ver- lag 1978.

18. Kops ER, Herzog H, Schmid A, Holte S, Feinendegen LE. Performance characteristics of an eight-ring whole body PET scanner. J Comput Assist Tomogr 1990; 14(3): 437-45.

19. Toorogian SA, Mulholland GK, Jewett DM, et al. Routine production of 2-deoxy-2-(’*F) fluoro -D-glucose by direct nucleophilic exchange on a Quaternary 4-aminopyridinium resin. Int J Rad Appl Instrum 1990; 17: 273-9.

20. Abumrad NN, Rabin D, Diamond MP, Lacy WW. Use of a heated superficial hand vein as an alternative site for the the measurement of amino acid concentrations and for the study of glucose and alanine kinetics in man. Metabolism

21. Bergstrom M, Eriksson L, Bohm C, Blomqvist G, Litton J. Correction for scattered radiation in a ring detector positron camera by integral transformations of the projections. J Comput Assist Tomogr 1983; 7: 42-50.

22. Maor MH, Maddux B, Osborne BM, et al. Stages IE and IIE non-Hodgkin’s lymphomas of the stomach. Comparison of treatment modalities. Cancer 1984; 54: 2330-7.

23. Taal BG, Burgers JMV, van Heerde P, Hart AAM, Somers R. The clinical spectrum and treatment of primary non- Hodgkin’s lymphoma of the stomach. Ann Oncol 1993; 4 (10): 839-46.

24. Fleming ID, Mitchel S, Dilawari RA. The role of surgery in the management of gastric lymphoma. Cancer 1982; 49: 1135-41.

1981; 30: 936-40.

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.

584 M. Rodriguez et al. Acta Oncologica 36 (1997)

25. Tio TL, den Hartog Jager FCA, Tytgat GNJ. Endoscopic ultrasonography in detection and staging of gastric non- Hodgkin lymphoma. Scand J Gastroenterol 1986; 21(Suppl

26. Kataoka M, Kawamura M, Tsuda T, et al. The role of gallium 67 imaging in non-Hodgkin’s lymphoma of the gas- trointestinal tract. Eur J Nucl Med 1990; 17: 142-7.

27. Douds HN, Berens SV, Long RF, Caplan GE. 67Ga-citrate scanning in gastrointestinal malignancies. Clin Nucl Med 1978; 3: 179-83.

28. Ichiya Y, Oshiumi Y, Kamoi I, et al. 67Ga scanning and

123): 52-8.

upper gastrointestinal series for gastric lymphoma. Radiology 1982; 142 187-92.

29. Que L, McCartney W, Hankins A, Hoffer P. Lymphomatous involvement of the stomach demonstrated by gallium-67 scan- ning. Am J Dig Dis 1974; 19: 271-4.

30. Eikman EA, Tenorio LE, Frank BA, et al. Gallium-67 accu- mulation in the stomach in patients with postoperative gastri- tis. J Nucl Med 1980; 21: 706-7.

31. MacMahon H, Vyborny C , Sephardari S, et al. Gallium accumulation in the stomach a frequent incidental finding. Clin NucI Med 1985; 10: 719-23.

Act

a O

ncol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

117.

167.

153.

80 o

n 05

/20/

14Fo

r pe

rson

al u

se o

nly.