—  SPECIALTY CONFERENCE  —

Gastrointestinal Pathology

Case 5 - Gastroduodenal Lymphocytic Phlebitis,
associated with Lymphocytic Gastritis and Gastric Antral Ulceration

Susan C. Abraham
Mayo Clinic
Rochester, MN


Click on each slide thumbnail image for an enlarged view
Clinical History
A 68-year-old man was admitted to an outside hospital because of epigastric pain, melena, and hematemesis that required multiple transfusions. His past medical history was significant for cardiomyopathy that had required pacemaker insertion, and he was receiving aspirin and Coumadin at the time of his gastrointestinal bleed. Upper endoscopic examination during that admission revealed a large gastric ulcer located in the posterior aspect of the antrum. The patient was told that he was negative for H. pylori (although it is not clear from the medical record how that determination was made). He was started on Prevacid 30 mg b.i.d Because of persistent mild epigastric pain as an outpatient, repeat upper endoscopy was performed 9 months later and showed a persistent posterior antral ulcer that measured 5 cm. He eventually underwent hemigastrectomy with vagotomy and a Billroth II gastroduodenal anastomosis. The slide shown includes both antral ulcer and adjoining viable antral mucosa and wall. The duodenum (to be shown later) had similar findings but lacked ulcers.


Case 5 - Figure 1 - Active chronic ulcer site in the antrum shows luminal fibrinoinflammatory exudate and underlying granulation tissue.
Case 5 - Figure 2 - Obliterative phlebitis is present in a submucosal vein beneath non-ulcerated gastric mucosa. The adjacent artery is normal.
Case 5 - Figure 3 - Phlebitis also involves vessels of the muscularis propria and subserosa, seen here in the duodenal wall.


Case 5 - Figure 4 - Prominent intra- and perivenular lymphoplasmacytic infiltrates.
Case 5 - Figure 5 - Granulomatous phlebitis.


Case 5 - Figure 6 - Lymphocytic gastritis in the non-ulcerated gastric antral mucosa.
Case 5 - Figure 7 - Mild lymphocytosis is also present in the duodenal epithelium, and there is associated villous blunting.

Follow-up
The patient initially did well, but at 16 months after surgery he began to complain of severe cramping epigastric pain, nausea, vomiting, intractable hiccupping, and heartburn. He was hospitalized 8 times over the subsequent 7 months. An extensive clinical and radiologic work-up included CT scans of the head, abdomen, and pelvis, and upper GI small bowel follow-through, which were negative. Doppler ultrasound of the mesenteric arteries and celiac axis showed patent vessels without significant stenoses. Upper endoscopy revealed diffuse bile reflux gastritis, as well as grade 2-3 esophagitis and hiatal hernia, but there was no recurrence of the gastric ulcer. A colonoscopy to the level of the terminal ileum showed no evidence for inflammatory bowel disease. Gastric biopsies (taken 25 months after surgery) showed a pattern of reactive gastropathy consistent with bile reflux in the setting of gastro-enteric anastomosis; lymphocytic gastritis was no longer present and there was negligible chronic inflammation within the lamina propria. The remainder of the biopsies from the duodenum, terminal ileum, and colon were normal.

Macroscopic Examination
The resection specimen included antrum, antral-fundic transition zone, and proximal duodenum. At the time of gross examination, a 4 cm x 3 cm red, hemorrhagic ulcer was noted in the distal antrum, whereas the remainder of the gastric mucosa was reported to be normal, without focal lesions. The duodenum contained a 0.4 cm mucosal polyp within the bulb (which proved to be heterotopic gastric fundic mucosa), but was otherwise grossly normal.

Microscopic Examination
Histologic sections revealed a benign chronic antral ulcer, with a granulation tissue base that extended into the submucosa. Lymphocytic and granulomatous phlebitis involved veins in the submucosa and muscularis propria; there was limited material for evaluation of the subserosal vessels, but one section containing subserosal adipose tissue also demonstrated lymphocytic phlebitis of a small subserosal vein. The phlebitic lesions were present in the veins deep to the antral ulcer as well as in the gastric and duodenal walls away from the ulcer site. Adjacent arteries and nerves were completely spared and appeared histologically normal. In some of the involved veins, vessel walls were infiltrated by small lymphocytes, which also formed cuffs in the perivenular tissue. In other veins, the inflammatory infiltrate included a prominent component of epithelioid histiocytes and rare multinucleated giant cells in addition to small lymphocytes; granulomas outside of vein walls were completely absent. Smaller numbers of plasma cells and large activated lymphocytes were also present. We saw no fibrinoid necrosis of the vein walls, and only a rare small venous thrombus was identified. However, the lumens of involved veins were variably compromised, including some veins with complete disruption of the wall and obliterative phlebitis.

The antral mucosa away from the ulcer site showed lymphocytic gastritis, with numerous small intraepithelial T-cells within the surface epithelium and foveolar neck epithelium, and numerous plasma cells within the superficial lamina propria. Both the resection specimen and a set of three previous gastric biopsies were negative for H. pylori. The duodenal bulb contained heterotopic gastric fundic mucosa as well as extensive gastric mucin cell metaplasia and near-complete blunting of the duodenal villi. Both the metaplastic gastric-type epithelium and the non-metaplastic intestinal epithelium also showed a lymphocytosis, which was much more prominent in the gastric-type epithelium (>100 lymphocytes/100 epithelial cells) than in the intestinal epithelium (20-40 lymphocytes/100 epithelial cells).

Diagnosis
Gastroduodenal lymphocytic phlebitis, associated with lymphocytic gastritis and gastric antral ulceration

Discussion
Among the vasculitic diseases that have been described to affect the gastrointestinal tract – e.g., polyarteritis nodosa, Churg-Strauss syndrome, Henoch-Schonlein purpura, hemolytic-uremic syndrome, Wegener's granulomatosis, enterocolic lymphocytic phlebitis, Behcet's disease, and collagen-vascular disease-associated vasculitis – only lymphocytic phlebitis exclusively affects the mural and mesenteric venous system.

In 1976, Stevens first reported this distinctive form of vasculitis in a previously healthy 36-year-old woman who underwent a right hemicolectomy for sudden, severe abdominal pain and diarrhea. [1] Histologically, there were numerous phlebitic lesions throughout the bowel wall, ranging from fibrinoid necrosis of vein walls to a more chronic, granulomatous perivenular infiltrate that resulted in partial to complete vein occlusions. Unexpectedly, the arteries and lymphatics were completely spared from this inflammatory reaction, which the authors speculated might be the result of an immunologic reaction to material absorbed from the bowel lumen. [1] Termed "necrotizing and giant cell granulomatous phlebitis" by the authors, this unusual form of vasculitis was not described again in the luminal GI tract for 13 years. In 1989, Saraga reported three patients who suffered from intestinal ischemic necrosis associated with lymphocytic phlebitis and proposed the term "enterocolic lymphocytic phlebitis." [2] It is this name (ELP) by which this distinctive type of primary gastrointestinal vasculitis is now known, although various authors have referred to similar cases as "granulomatous giant cell polyphlebitis," [3] "mesenteric inflammatory veno-occlusive disease," [4, 5, 6] "chronic intestinal lymphocytic microphlebitis," [7] "isolated granulomatous phlebitis," [8] and "intramural mesenteric venulitis." [9]

ELP is rarely diagnosed, and although almost 40 patients have now been reported in the English literature, most are in the form of case reports [1, 3, 5, 8, 9, 10, 11, 12, 13] or small case series. [7, 15] The three largest series, by Burke, [14] Flaherty, [4] and Saraga, [15] have respectively described 10, 7, and 6 patients with ELP. Several distinctive clinicopathologic characteristics have emerged from these reports. Most patients with ELP present with acute abdominal symptoms including severe abdominal pain, diarrhea, and/or bleeding. [1, 4, 8, 9, 13, 14, 15] Patients with ELP have ranged from 27 – 78 years and there has been no clear gender predilection. [4, 14, 15] Most cases of ELP have involved the colon (particularly the right side) [1, 4, 5, 7, 8, 9, 10, 12, 13, 14, 15] but cases involving the ileum [3, 4, 14, 15] or both small bowel and colon [4, 6, 15] have also been described; rarely, there is pancolonic necrosis. [8] Characteristically, these patients recover quickly following resection of the involved bowel, without intestinal or systemic recurrence of the vasculitis.

The patient presented here with lymphocytic phlebitis is unusual in two respects. First, his symptoms followed a more chronic course than is typical, since his epigastric pain and gastric ulceration were documented for a total of 9 months prior to surgical intervention. However, several patients with ELP affecting the lower gastrointestinal tract have also been noted to have unusual, subacute, or chronic presentations. In the series by Saraga, 4 of 6 patients suffered from prodromal abdominal pain, weight loss, and/or rectal bleeding occurring from several weeks to months before the onset of acute abdominal signs. [15] Two patients demonstrated histologic changes indicative of chronic ischemia in their resection specimens. One of these patients in particular, a 70-year-old man, suffered from abdominal pain for years and rectal bleeding for several months preceding surgery. [15] Furthermore, 4 patients in that series also complained of recurrent abdominal pain after surgery, although recurrent ELP was histologically documented in only one of these cases, and no patient required re-operation. [15] This suggests that ELP may not always be an acute illness, and may not always fully resolve with surgical intervention. Although our patient did suffer from abdominal complaints after his surgery, the fact that this occurred fully 16 months after an initially uneventful recovery, as well as the fact that no recurrent gastric ulceration occurred, likely indicates that his complaints were not the result of recurrent lymphocytic phlebitis.

Second, the lymphocytic phlebitis in our patient involved the stomach and duodenum, without evidence for ileal or colonic ischemia. Involvement of the upper GI tract by lymphocytic phlebitis has not been previously reported. However, lymphocytic/granulomatous phlebitis has been reported in other extracolonic sites including the pulmonary veins, [16] liver, [17] skin (sometimes affecting both the cutaneous and mesenteric veins), [18] gallbladder, [14] and omentum. [14] Therefore, there is no theoretical reason why a similar process should not also affect the gastric and/or duodenal venous system.

Aside from these differences in clinicopathologic presentation, the histologic findings in our patient were otherwise typical for ELP. There was extensive phlebitis of the small and medium-sized submucosal, mural, and perigastric veins, whereas the adjacent arteries were completely spared. The vein walls and perivenular areas were infiltrated by cuffs of small, mature lymphoid cells which were admixed with scattered plasma cells and eosinophils. Tuppy et al have previously shown that the many of the perivenular T-cells are CD4-/CD8+/TIA-1+ cytotoxic T-cells, [13] and our case confirmed this immunophenotypic profile. In some areas, there was also prominent granulomatous perivenular inflammation. Giant cell phlebitis was reported in 30% of Burke's cases of ELP, [14] as well as those reported by Stevens, [1] Flaherty, [4] Leu, [3] Satge, [6] Martinet, [8] and Saraga. [15] Although some authors have described thromboses and fibrinoid necrosis of the vein walls, [1, 4, 5, 8, 12, 15] we found only rare small thrombi and no fibrinoid necrosis. However, many of the vein lumens were markedly narrowed or obliterated by the inflammatory reaction. Notably, the phlebitis in our patient was present not only in association with the gastric antral ulceration but also within non-ischemic areas of the stomach and duodenum – a phenomenon that has been well-described in some cases of ELP from the lower GI tract. [7, 15]

The pathogenesis of lymphocytic phlebitis is unclear. In their initial report, Stevens proposed that ELP might represent an immunologic reaction to material absorbed from the bowel, [1] and Saraga likewise suggested a hypersensitivity-type reaction. [2] Many of the reported patients with ELP were receiving one or multiple types of medications, [2, 11, 14, 15] but this is not a universal feature [9] and no one drug has emerged as a consistent etiologic agent. The fact that our patient showed both lymphocytic gastritis and duodenitis which had completely resolved (without dietary therapy and without other clinical evidence for celiac disease) 2 years after his surgical resection would also suggest the possibility of a drug hypersensitivity reaction, since some cases of lymphocytic colitis are drug-associated. [19, 20] However, a detailed list of medications taken by the patient at the time of his presentation and outside clinical workup was not available to us.

Why has lymphocytic phlebitis not been previously described to affect the upper gastrointestinal tract? First, processes that result in classic histologic features of ischemia in the colon may not be recognizable as such when they affect the stomach. For example, administration of Kayexalate in hypertonic sorbitol, which causes acute ileal/colonic ischemic necrosis in a small subset of patients, typically shows only nonspecific mucosal erosions or ulcerations when the upper gastrointestinal tract is involved. Similarly, Burke et al described a non-healing gastric ulcer that resulted from polyarteritis. [14] The diagnosis of lymphocytic phlebitis requires not only a level of awareness on the part of the pathologist, but also requires deep submucosal or mural veins for examination – material which is unlikely to be obtained except through surgical resection. As surgical resection for benign gastric ulcers is not typical first-line therapy, we suggest that gastric involvement by lymphocytic phlebitis may not be as rare as this single case would imply.

References

  1. Stevens SM, Gue S, Finckh ES. Necrotizing and giant cell granulomatous phlebitis of caecum and ascending colon. Pathology 1976;8:259-64.
  2. Saraga EP, Costa J. Idiopathic entero-colic lymphocytic phlebitis. A cause of ischemic intestinal necrosis. Am J Surg Pathol 1989;13:303-8.
  3. Leu HJ, Brinninger G, Pernegger C, Odermatt B. Primary granulomatous giant cell polyphlebitis of visceral veins. Virchows Arch A Pathol Anat Histopathol 1993;423:519-21.
  4. Flaherty MJ, Lie JT, Haggitt RC. Mesenteric inflammatory veno-occlusive disease. A seldom recognized cause of intestinal ischemia. Am J Surg Pathol 1994;18:779-84.
  5. Rademaker J. Veno-occlusive disease of the colon – CT findings. Eur Radiol 1998;8:1420-1.
  6. Satge D, Jardel P, Lavoine E, Goburdhun J, Flejou J. Fatal mesenteric ischemic accident caused by mesenteric inflammatory veno-occlusive disease. Ann Pathol 1999;19:525-8.
  7. Endes P, Molnar P. Chronic intestinal lymphocytic microphlebitis. Acta Morphol Hung 1992;40:137-47.
  8. Martinet O, Reis ED, Joseph JM, Saraga E, Gillet TM. Isolated granulomatous phlebitis: rare cause of ischemic necrosis of the colon: report of a case. Dis Colon Rectum 2000;43:1601-3.
  9. Corsi A, Ribaldi S, Coletti M, Bosman C. Intramural mesenteric venulitis. A new cause of intestinal ischaemia. Virchows Arch 1995;427:65-9.
  10. Arain FA, Willey J, Richter J, Senagore A, Petras R. An unusual presentation of enterocolic lymphocytic phlebitis. J Clin Gastroenterol 2002;34:252-4.
  11. Chergui MH, Vandeperre J, Van Eeckhout P. Enterocolic lymphocytic phlebitis: a case report. Acta Chir Belg 1997;97:293-6.
  12. Haber MM, Burrell M, West AB. Enterocolic lymphocytic phlebitis. Clinical, radiologic, and pathologic features. J Clin Gastroenterol 1993;17:327-32.
  13. Tuppy H, Haidenthaler A, Schandalik R, Oberhuber G. Idiopathic enterocolic lymphocytic phlebitis: a rare cause of ischemic colitis. Mod Pathol 2000;13:897-9.
  14. Burke AP, Sobin LH, Virmani R. Localized vasculitis of the gastrointestinal tract. Am J Surg Pathol 1995;19:338-49.
  15. Saraga E, Bouzourenne H. Enterocolic (lymphocytic) phlebitis: a rare cause of intestinal ischemic necrosis: a series of six patients and review of the literature. Am J Surg Pathol 2000;24:824-9.
  16. Crissman JD, Koss M, Carson RP. Pulmonary veno-occlusive disease secondary to granulomatous venulitis. Am J Surg Pathol 1980;4:93-9.
  17. Nakanuma Y, Ohta G, Doishita K, Maki H. Granulomatous liver disease in the small hepatic and portal veins. Arch Pathol Lab Med 1980;104:456-8.
  18. Lie JT. Primary cutaneous granulomatous phlebitis with visceral involvement: unmasking of a masked villain. Mod Pathol 1996;9:719-24.
  19. Baert F, Wouters K, D'Haens G, Hoang P, Naegels S, D'Heygere F, Holvoet J, Louis E, Devos M, Geboes K. Lymphocytic colitis: a distinct clinical entity? A clinicopathological confrontation of lymphocytic and collagenous colitis. Gut 1999;45:375-81.
  20. Wang N, Dumot JA, Achkar E, Easley KA, Petras RE, Goldblum JR. Colonic epithelial lymphocytosis without a thickened subepithelial collagen table: a clinicopathologic study of 40 cases supporting a heterogeneous entity. Am J Surg Pathol 1999;23:1068-74.