Clinical Radiology (1981) 32, 63-71 0009-9260/81/01530063502.00 © 1981 Royal College of Radiologists

Primary Lymphoma of the Gastrointestinal Tract OSCAR CRAIG and ROGER GREGSON

Department o f Diagnostic Radiology, St Mary's Hospital, London

Twenty-two cases of primary lymphoma of the gastrointesinal tract are reported. These have been seen over the past 16 years and consist of seven gastric, 13 small bowel and two caecal lesions. This number represents 0.07% of the barium examinations of the stomach and small bowel, conducted over this period. All the cases were examined by the conventional barium meal or barium meal and follow-through examinations using non- flocculable barium. All the lymphomas were verified by histological examination. The cases are analysed as to clinical and radiological presentation, and the clinical and radiological diagnostic scores are assessed. There was a very poor diagnostic score on clinical grounds alone; a high diagnostic score for radiological examination of the stomach of 71%, contrasted with a low radiological score of 26% for the small bowel. The radiological features of lymphoma in the stomach and small bowel are described and the differential diagnosis discussed. To improve the diagnostic score on radiological examination, an awareness of gastrointestinal lymphoma is essential, as is a knowledge of the possible radiological changes. The barium examination should carefully search for these changes where lymphoma is clinically possible.

Gall and Mallory (1942) state that because lymphoma is common- in the retroperitonal lymph nodes, involvement of the gastrointestinal tract by lym- phoma in disseminated cases is as high as 10-20%. Bush and Ash (1969) give a figure of 3% for involve- ment of the small intestine in disseminated disease and Rosenberg et al. (1961) found the stomach was involved in 3.5% of disseminated disease. From a series of 1555 cases of lymphoma, Bush and Ash (1969) found 2.5% had a primary tymphoma in the stomach and 2.1% had a primary lymphoma in the small bowel.

Dawson et al. (1961) make the point that very few large series of the primary lymphomas of the gastro- intestinal tract have been published, and no centre can see sufficient of these to be able to lay down use- ful criteria. Faulkner and Dockerty (1952) reported 33 cases, Allen et al. (1954) 34 cases, Raiford (1933) 32 cases and Warren (1959) 24 cases.

Primary intestinal lymphoma is rare and it was thought worthwhile to review the cases seen in the Department of Diagnostic Radiology at St Mary's Hospital over the past 16 years.

PATIENTS AND METHODS

Twenty-two cases of primary gastrointestinal lymphoma are recorded over this period. This repre- sents an incidence of 0.07% of all the barium studies of the stomach and small bowel performed over the 16 years. Fifteen of the cases were seen in the last 10 years.

The criteria used for considering the lesion to be a primary lymphoma were those laid down by Dawson etal. (1961).

1. When the patient was first seen there were no palpable enlarged lymph nodes.

2. Chest radiographs showed no obvious enlarge- ment of the mediastinal glands.

3. The white cell count, total and differential, were within normal limits.

4. At laparotomy, the bowel lesion predomi- nated, the only lymph nodes affected being those in the immediate neighbourhood.

5. The liver and spleen appeared free of tumour in every case.

All the cases in this paper were examined radio- logically by the conventional barium meal or barium meal and follow-through examinations. All the diag- noses were confirmed by biopsy either at endoscopy or operation.

Pathology

Classification of the varieties of lymphoma presents a difficulty. Numerous pathological classifi- cations have been suggested for lymphoma since these cases have been diagnosed. For clarity the classifi- cation used by the pathologists at the time of diag- nosis will be used throughout the paper. The lym- phomas in this series then fall into the following groups, reticulum cell sarcoma, lymphosarcoma, malignant lymphoma and Hodgkin's disease. Henry

64 CLINICAL RADIOLOGY

and Farrer-Brown (1977) in an editorial in the Lance t are quoted as saying that they would question the diagnosis of primary Hodgkin's disease of the gastro- intestinal tract. Using Rappaport ' s (1966) classifi- cation, these lymphomas would be described as mixed lymphocyt ic and hist iocytic types, poorly differentiated lymphocyt ic , etc. Some modern patho- logical classifications include plasmacytoma of the gastrointestinal tract and this would now be considered in the lymphoma group. There were in fact two cases of gastric plasmacytoma, seen during this period, but these have not been included in the series.

RESULTS

Table 1 lists the patients giving sex, age, history, site and length o f history at the time of diagnosis.

There were five cases with a history of less than one month, three of these were gastric lymphoma, one ileal and one caecal. The ileal lymphoma presented with an obstruction and the caecal lesion presented as a perforation. The longest history of six years was in Case 2, which presented as a mass in the right iliac fossa. Dawson et al. (1961) state that primary intestinal lymphomas are approximately twice as common in men as in women. In this present

series this has not been the case, 11 being male and 11 female.

Ngan and James (1973) say that primary intestinal lymphoma is much commoner in the earlier years of life, but only eight o f the lymphomas in this series were less than 50 years o f age, 14 were over 50 and 10 of these were over 60.

The discrepancy between the number of cases and the sites in which they were found was due to the spread of the lesion, i.e. two lesions involved both the ileum and caecum (Cases 12 and 20) and two lesions involving the stomach had spread into the duodenum (Cases 13 and 22). There were no cases in the oesophagus and none in the large bowel beyond the caecum.

The stomach is thought to be involved b y lym- phoma more frequently than the small bowel and Carnovale et al. (1977) state that the stomach, small bowel and large bowel are involved in decreasing frequency. However, in the present series seven cases originated in the stomach and 13 cases in the small bowel.

In the paper by Dawson et al. (1961) it is stated that there was no tendency for any histological tumour to be associated with any particular macro- scopic appearance or to show any predilection for any one site. This was not the case in the present

Table 1 - Details of 22 patients with primary gastrointestinal lymphoma

Case Sex Age Histology Site Length of no. history

1 M . 78 ML Stomach 1 year 2 F 36 RCS Jejunum 6 years 3 F 60 RCS Jejunum 3 months 4 M 60 HD Jejunum 4 years 5 F 70 RCS Stomach 2 years 6 M 47 LS Ileum 2 months 7 M 58 RCS Jejunttm 5 years 8 M 78 ML Stomach 2 months 9 F 61 RCS Caecum 6 months

10 F 47 RCS Ileum 1 week 11 F 59 LS Ileum 6 weeks 12 F 29 RCS Ileum and caecum 2 months t 3 M 34 ML Stomach and duodenum 6 months 14 M 49 RCS Jejunum ? 15 M 53 ML Stomach 2 weeks 16 F 70 LS Jejunum ? 17 M 62 LS Caecum 6 days 18 F 61 RCS Jejunum 5 months 19 F 77 LS Stomach 2 weeks 20 M 33 LS Ileum and caecum 2 years 21 F 35 HD Jejunum 1 year 22 M 53 ML Stomach and duodenum 3 weeks

RCS, Reticulum cell sarcoma; LS, lymphosarcoma; ML, malignant lymphoma; HD, Hodgkin's disease.

L Y M P H O M A O F T H E G I T R A C T 65

series. Five of the nine cases classified as reticulum cell sarcoma arose in the jejunum, and all the cases classified as malignant lymphoma, i.e five, arose in the stomach.

Clinical Presentation

The cases presented with a very varied clinical picture which included a palpable mass, epigastric pain, lower abdominal pain, diarrhoea, intestinal obstruction, anorexia, weight loss, haematemesis and melaena, and perforation. Table 2 lists the incidence of these.

Table 2 - Clinical presentation (all cases)

Mass 6 Epigastric pain 5 Vomiting 4 Lower abdominal pain 11 Diarrhoea 7 Intestinal obstruction 1 Anorexia 4 Weight loss 10 Perforation 3 Haematemesis 1 Melaena 2

In this series the most common symptoms of gastrointestinal lymphoma were lower abdominal pain, diarrhoea and weight loss. A mass was clinic- ally palpable in six cases. It is noted that three cases presented as a perforation clinically. These cases were a Hodgkin's disease of the jejunum, a reticulum cell sarcoma of tl~e ileum and lymphosarcoma of the caecum. All had gas under the diaphragm and pro- ceeded to laparotomy. Ngan and James (1973) say intestinal obstruction is a rare presentation of gastro- intestinal lymphoma. Marshak e t al. (1961) say lymphosarcoma may occasionally produce intestinal obstruction, if accompanied by mucosal ulceration and secondary infection. There was only one case in the present series that presented as an intestinal obstruction and that was a reticulum cell sarcoma of the ileum.

Table 3 - Clinical presentation of gastric lymphoma

Pain 5 Mass 0 Haematemesis 1 Melaena 1 Vomiting 3 Anorexia 3 Weight loss 3

The clinical presentation of gastric lymphomas was analysed separately and Table 3 lists the frequency of each symptom.

Pain was the most frequent complaint, followed by anorexia, vomiting and weight loss with equal fre- quency. Even though Ngan and James (1973) say that a palpable mass is a not uncommon presentation of gastric lymphoma, there were no cases presenting clinically with a gastric mass in this series.

Radiology

Primary lymphoma may present radiologically as a mass lesion, stricture, mucosal infiltration, abscess, fistula, obstruction or perforation. The cases in this series were analysed as to the frequency of the pre- dominant radiological changes, and the results are noted in Table 4. The majority of lymphomas pre- sented more than one radiological change.

Table 4 - Radiological findings in all cases

Ulceration 3 Stricture 3 Mass 6 Perforation 4 Mucosal infiltration 9 Abscess 1 Fistula 2 Ob stru ction 1

The most common radiological appearances in the small bowel were irregular mucosal infiltration and a mass lesion (Figs 1, 2). Although only three cases presented as a perforation clinically, a further case was diagnosed radiologically as having perforated locally. A small bowel abscess (Fig. 3) was a rare presentation. Figs 4 and 5 are examples of stricture formation. Ulceration may occur in the small bowel and Fig. 6 shows a case with multiple small bowel ulcers confirmed at operation. There were two cases with fistulae and Fig. 7 shows a case where an external fistula post-operatively communicated with the duodenum and the large bowel. In the majority of cases there was more than one radiological sign. Figs 8 and 9 show examples of mucosal irregularity and narrowing in lower ileal coils which appear matted together, and a filling defect in the caecum. Some cases radiologically had a pattern suggesting malabsorption, but the diagnosis of idiopathic steatorrhoea was not made as there were features suggesting the underlying cause. Fig. 10 shows a pattern of thickened mucosal folds and dilatation in

66 C L I N I C A L R A D I O L O G Y

Fig. 1 - Distortion and irregularity of the mucosa over a small segment of jejunum (arrow). Reticulum cell sarcoma.

Fig. 2 - A mass outlined by markers. Irregularity of the ileum due to reticulum cell sarcoma.

a pat ient wi th diarrhoea. The s tomach in this case was irregular and inf i l t rated suggesting lymphoma. In some o f the cases the mucosal inf i l t ra t ion was widespread and involved the d u o d e n u m in addi t ion to the j e junum. The de fo rmi ty o f the duodenum is well seen in Fig. 11.

The gastric lesions were analysed separately and

Fig. 3 - Large abscess (arrow) arising from a jejunal reti- cnlum cell sarcoma. Fig. 4 - Jejunal strictures (arrows) due to lymphosarcoma.

LYMPHOMA OF THE GI TRACT 67

Fig. 5 - Jejunal strictures due to reticulum ceU sarcoma.

Fig. 7 - Fistulae into duodenum and large bowel (arrows). Reticulum cell sarcoma.

three important radiological changes were noted; a radiological mass was seen in three cases, mucosal hypertrophy or prominence in five cases and ulcera- tion in two. There were, as elsewhere in the gastro- intestinal tract, more than one set of changes in each case. Fig. 12 shows a coned view of a lesser curve ulcer with diffuse mucosal changes. One of the findings in the stomach, which strongly suggests lymphoma, is where mucosal involvement continues across the pylorus into the duodenum (Ngan and James, 1973; Sutton and Grainger, 1975). Fig. 13 shows a lesion in the pyloric antrum extending across the pylorus into the duodenum.

Radiological and Clinical Diagnostic Score

The accuracy of the clinical diagnosis recorded in the patients' notes was assessed for the stomach, small bowel and caecum. The result is given in Table 5.

There were no pre-investigation correct diagnoses. In the stomach two were diagnosed as peptic ulcers and two as carcinomas. Three cases were merely listed for investigation because of gastrointestinal symptoms. Although a correct diagnosis was not

Fig. 6 - Multiple small bowel ulcers (arrows) due to Hodgkin's disease.

Table 5 - Clinical diagnosis (22 patients)

Site Definite Suggested Wrong No definite diagnosis

Stomach - - 4 3 Small bowel - - 5 8 Caecum - - - 2

68 CLINICAL RADIOLOGY

Fig. 8 - Mucosal irregularity in coils of adherent ileum (arrow). Lymphosarcoma.

Fig. 9 - Irregularity of caecum and mucosal 'mass' project- ing into lumen (arrow). Lymphosarcoma.

made on clinical grounds alone, this is no t in tended to imply that investigations such as endoscopy or l aparo tomy would no t have led to the correct diagnosis even in the absence of radiological investi- gation. In the case of the small bowel , the five incorrect diagnoses comprised two diagnosed as Crohn's disease, one as intest inal tuberculosis , one as benign je junal ulcerat ion and one as idiopathic steatOrrtioea. When no definitive diagnosis was made clinically, the picture was perforat ion or obst ruct ion or mass. One of the caecal lesions presented as a perforat ion and the other was listed as a mass for investigation. In none of the cases was l ymphoma considered as a clinical diagnosis.

The accuracy of the radiological diagnosis was assessed from the radiological reports and Table 6 lists the results.

The striking finding was the high diagnostic score in the stomach, i.e. five correct out of seven cases. One case was incorrect ly diagnosed as a carc inoma

Table 6 - Radiological diagnosis (22 patients)

Site Definite Suggested Wrong No definite diagnosis

Stomach 5 1 1 - Small bowel 4 2 5 2 Caecum - - 1 1 Fig. 10 - Malabsorption like pattern in jejunum and irregular

infiltration of stomach. Malignant lymphoma.

L Y M P H O M A O F T H E G I T R A C T 69

Fig. 11 - Irregular infiltration of duodenal mucosa. Reticulum cell sarcoma.

Fig. 12 - Ulcer on lesser curve of stomach (arrow) with surrounding extensive mucosal thickening and int~fltration. Lympho- sarcoma.

Fig. 13 - Mass lesion in the pyloric antrum extending across into the duodenum (arrows). Malignant lymphoma.

and in another case carcinoma was thought likely but lymphoma considered possible. In the small bowel there were four correct diagnoses out of 13 cases. Five incorrect diagnoses were made, and these included tuberculosis, Crohn's disease, Zo l l inger - Ellison syndrome, a case diagnosed merely as mal- absorption and one diagnosed as small bowel obstruc- tion. It could be argued that these two cases should be included in the no definitive diagnosis group. Two of the small bowel lesions only had plain films as they presented as intestinal perforations and these are included in the no diagnosis group. There were two caecal lesions, one presented as a perforation and this was the radiological diagnosis on plain films. The other case was incorrectly diagnosed as a carcinoma.

Although the total figure of cases is small, i.e. 22 cases, the radiological accuracy for the whole group of 41% compares favourably with a figure of 30% quoted by Welborn e t al. (1967). The diag- nostic accuracy for the small bowel and caecum in this series is 26%. A diagnostic accuracy of 71% is seen in the gastric lymphoma group.

70 C L I N I C A L R A D I O L O G Y

Table 7 - Survival (22 patients)

<6 month 6 m o n t h s - 2 - 5 5year s + Los t to 2 years years fol low-up

5 1 3 8 5

Prognosis

Table 7 gives the survival rate for the 22 cases. Five cases survived less than six months and eight cases survived more than five years. If the cases lost to follow-up are excluded then 30% died within six months and 50% were alive after five years. Of the eight surviving over five years, one is alive at seven years (Case 12), one at 10 years (Case 1) and another at 14 years (Case 6).

Although secondary involvement of the gastro- intestinal tract by disseminated lymphoma is invari- ably associated with a fatal outlook, this is not the case for primary lymphoma. Bush and Ash (1969) reported a two-year recurrence-free rate of 50% and Fu and Perzin (1972) reported a five year survival rate of 40%. The figure of a five-year survival rate of around 50% in this series compares favourably with these figures. It would appear that the best results arise from widespread resection followed by radio- therapy (Ngan and James, 1973).

DISCUSSION

The high diagnostic score for radiology in the case of the gastric lesions is easy to explain, as at least two of the features of lymphoma were recognised in each of the five correct cases. The most frequent findings were mucosal hypertrophy or prominence associated with mucosal ulceration, and a mass extending across the pylorus into the duodenum.

The features described as suggestive of gastric lymphoma are:

1. Diffuse mucosal hypertrophy, irregularity or thickening.

2. Multiple ulcerations. This feature was not pre- sent in the present series, but is well described (Bloch, 1967; Highman and Key, 1962; Privett e t al., 1977).

3. A single ulcer associated with diffuse mucosal thickening.

4. A mass lesion or a mucosal irregularity extend- ing across the pylorus into the duodenum.

These features are referred to by Ngan and James (1973) and Sutton and Grainger (1975) amongst others. In the radiological reports of the present series no mention is made of the pliability of the

gastric wall. Privett e t al. (1977) say that a feature suggestive of gastric lymphoma is distensibility of the wall of the stomach in the presence of a mass. Sutton and Grainger (1975) mention lack of pliability of the gastric wall as a feature of reticulosis. Ngan and James (1973) mention reduced distensibility, with peristalsis impaired, as suggestive of gastric lymphoma. Reports in the literature vary sufficiently to make this feature relatively unimportant in arriving at a diagnosis of lymphoma. The main radiological differential diagnosis in gastric lymphoma will be from simple ulceration, aphthous ulceration, carci- noma and pseudo-lymphoma (Martel e t al., 1976; Chiles and Platz, 1975). Pseudolymphoma is a benign proliferation of lymphoid tissue which may be a reaction to peptic ulceration, and the appearances can be mistaken for lymphoma both histologically and radiologically.

The diagnostic score for radiological investigation of the small bowel was only 26%. This compares with the figure of 30% quoted by Welborn e t al. (1967). Ngan and James (1973) say that despite some rather characteristic radiographic appearances, lymphoma of the small bowel is not often diagnosed pre-operatively. Cupps e t al. (1969) reviewed 46 cases of primary malignant lymphoma of the small bowel and 65% of the cases were indistinguishable radiologically from other neoplastic and non-neoplastic diseases of the small bowel. The main difficulty in the diagnosis is to distinguish primary lymphoma from regional enteritis and tuberculosis. Intestinal lymphangiec- tasis may also mimic lymphoma as may Cronkite- Canada syndrome. Lymphoma may present radiologically as a malabsorption pattern of a non- specific nature and the diagnosis of idiopathic steator- rhea may be made. It is of the utmost importance in these cases, using non-flocculable barium, to define the mucosa adequately, and to search for any mucosal irregularity or stricture that would suggest lymphoma.

Three of the cases in this paper had a history of coeliac disease. There is well-documented evidence of an association between primary lymphoma and coeliac disease (Barry and Read, 1973; Gough e t al. 1962; Loehr e t al., 1969). Such an association has also been described between lymphoma and Crohn's disease and even ulcerative colitis. Clinically, if a gluten sensitive enteropathy fails to respond to a gluten-free diet or relapses while on a gluten-free diet, this suggests the possibility of lymphoma. In certain geographical areas, namely Middle Eastern countries, the Mediterranean area, Central America and South Africa, there is a primary upper small intestine lymphoma (PUSIL), which must be con- sidered in the differential diagnosis of patients

LYMPHOMA OF THE GI TRACT 71

presenting with diarrhoea (Nasr et al., 1976; Tabbane et al., 1976).

Two rare conditions in the differential diagnosis are eosinophlic gastroenteritis and ulcerative non- granulomatous jejunitis.

The localised form of eosinophilic gastroenteritis may present as a pyloric mass indistinguishable from carcinoma or lymphoma. The diffuse form of eosinophilic gastroenteritis affects the duodenum and small bowel and produces extensive mucosal irregularity and even stricture formation. The changes

m a y be intermittent and undergo spontaneous remission. There may on occasion be an allergic history and a blood eosinophilia (Brostoff et al., 1970). It is frequently necessary to make the diagnosis by Crosby capsule biopsy. Ulcerative non- granulomatous jejunitis is an uncommon condition, where ulceration and even stricture formation may occur. Some authors have claimed it is a compli- cation of the coeliac disease. The relationship between lymphoma and ulcerative non-granulo- matous jejunitis is not clear, and cases diagnosed initially as ulcerative jejunitis have returned later with lymphoma (Freeman et al., 1977).

The radiological features of primary lymphoma seen in the series reported in this paper include stric- ture formation, masses, single or multiple ulceration, abscess formation and perforation. Aneurysmal dilatation of the bowel, said to be highly suggestive of lymphoma, was not a feature of any of these cases, nor was the multiple small polypoid type of infil- tration. Although the diagnosis is not made frequently, it is important to consider lymphoma when any of these radiological features are present. The conventional barium meal and barium meal and follow-through examinations with non-flocculable barium are satisfactory methods for the examination of the stomach and the small bowel. However, films of the small bowel must be taken frequently, prefer- ably at half-hourly intervals until the column reaches the caecum. These films should be examined by the radiologist and any doubtful areas screened. It is also necessary to screen the terminal ileum.

Although the diagnostic score radiologically is high for the stomach (71%), it is low for the small bowel (26%). If this diagnostic score is to increase, then radiologists examining the gastrointestinal tract must constantly be aware of the possibility of lym- phoma and of its varied radiological presentation.

Acknowledgements. The authors acknowledge all the clinical colleagues of St Mary's Hospital who kindly referred their cases for examination. We wish to express our thanks to Dr A. Boylston who re-examined all the biopsy material and made us aware of the modern pathological classification.

We are indebted to Dr Peter Cardew and the staff of the Audio-visual Department of St Mary's Hospital Medical School for reproducing the radiographs, and to Miss Siobhan Craig for typing and secretarial assistance.

REFERENCES

Allen, A. W., Donaldson, G., Sniffen, R. C. & Goodale, F. (1954). Primary malignant lymphoma of the gastro- intestinal tract. Annals of Surgery, 140, 428-437.

Barry, R. E. & Read, A. E. (1973). Coeliac disease and malig- nancy. Quarterly Journal of Medicine (New Series XLII), 168, 665-675.

Bloch, C. (1967). Roentgen features of Hodgkin's disease of the stomach. American Journal of Roentgenology, 99, 175-181.

Brostoff, J., Craig, O. & Everest, M. (1970). Eosinphilic gastroenteritis. British Journal of Surgery, 57, 653- 657.

Bush, R. A. & Ash, C. L. (1969). Primary lymphomas of the gastrointestinal tract. Radiology, 92, 1349-1352.

Carnovale, R. L., Goldstein, H. M., Zomoza, J. & Dodd, G. D. (1977). Radiologic manifestations of oesophageal lymphoma. American Journal o f Roentgenology, 128, 751-754.

Chiles, J. T. & Platz, L. E. (1975). The radiographic mani- festations of pseudo-lymphoma of the stomach. Radio- logy, 116,551-556.

Cupps, R, E., Hodgson, J. R., Docherty, M. B. & Adson, M. A. (1969). Primary lymphoma in the small intestine; problems of roentgenologic diagnosis. Radiology, 92, 1355-1362.

Dawson, 1. M. P., Comes, J. S. & Morson, B. C. (1961). Primary malignant lymphoid tumours of the intestinal tract. British Journal o f Surgery, 49, 80-89.

Faulkner, J. W. & Docherty, M. B. (1952). Lymphosarcoma of the small intestine. Surgery, Gynaecology and Obstetrics, 95, 76-84.

Freeman, H. J., Weinstein, W. M., Shnitka, T. K., Piercey. R. J. & Wensel, R. H. (1977). Primary abdominal lym- phoma~ presenting conditions of coeliac sprue or compli- cating dermatitis herpetiformis. American Journal of Medicine, 63,585-594.

Fu, Y. S. & Perzin, K. H. (1972). Lymphosarcoma of the small intestine. A clinicopathological study. Cancer, 29, 645-659.

Gall, E. A. & Mallory, T. B. (1942). Malignant lymphoma; clinico-pathological survey of 618 cases. American Journal of Pathology, 18, 381-429.

Gough, K. R., Read, A. E. & Naish, J. M. (1962). Intestinal reticulosis as a complication of idiopathic steatorrhoea. Gut, 3,232 239.

Henry, K. & Farrer-Brown, G. (1977). Primary lymphomas of the gastrointestinal tract and plasma cell tumours. Histopathology, 1, 53-76.

Highman, J. H. & Key, J. J. (1962). Multiple ulceration of the stomach in reticulum cell sarcoma. British Journal of Radiology, 35,614-618.

Loehr, W. J., Mujahed, Z. & Zahn, D. F. (1969). Primary lymphoma of the gastrointestinal tract; review of 100 cases. Annals of Surgery, 170, 232-238.

Marshak, R. H., Wolf, B. S. & Eliasoph, J. (1961). The roentgen findings in lymphosarcoma of the small intestine. American Journal of Roentgenology, 86, 682 692.