Article Text

Download PDFPDF

Endoscopic evaluation for cytomegalovirus enterocolitis after allogeneic haematopoietic stem cell transplantation
  1. Y Kakugawa1,
  2. M Kami2,
  3. T Kozu3,
  4. N Kobayashi3,
  5. H Shoda3,
  6. T Matsuda3,
  7. Y Saito3,
  8. I Oda3,
  9. T Gotoda3,
  10. S Mori4,
  11. R Tanosaki4,
  12. N Murashige4,
  13. T Hamaki4,
  14. S Mineishi4,
  15. Y Takaue4,
  16. T Shimoda5,
  17. D Saito6
  1. 1Endoscopy Division, the National Cancer Centre Hospital, Tokyo, Japan
  2. 2Haematopoietic Stem Cell Transplantation Unit, the National Cancer Centre Hospital, Tokyo, Japan
  3. 3Endoscopy Division, the National Cancer Centre Hospital, Tokyo, Japan
  4. 4Haematopoietic Stem Cell Transplantation Unit, the National Cancer Centre Hospital, Tokyo, Japan
  5. 5Pathological Division, the National Cancer Centre Hospital, Tokyo, Japan
  6. 6Endoscopy Division, the National Cancer Centre Hospital, Tokyo, Japan
  1. Correspondence to:
    Dr Y Kakugawa
    Endoscopy Division, the National Cancer Centre Hospital, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan; yakakuga{at}ncc.go.jp

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Cytomegalovirus (CMV) disease and graft versus host disease (GVHD) are serious complications after allogeneic haematopoietic stem cell transplantation (allo-SCT).1 As clinical manifestations overlap,2 it is difficult to make an early and accurate diagnosis of CMV enterocolitis in patients with gastrointestinal GVHD. This study aimed to estimate the usefulness of endoscopic examination for CMV enterocolitis after allo-SCT.

Between 1999 and 2003, 425 patients received allo-SCT at the National Cancer Centre Hospital. Eighty seven patients with gastrointestinal symptoms underwent colonoscopy, and gastrointestinal GVHD was diagnosed in 75 patients. Thirty three of these patients with a median age of 46 years (range 4–67) who had persistent diarrhoea and/or abdominal pain despite optimal treatment underwent repeat colonoscopy. After informed consent was obtained, we examined from the terminal ileum to the rectum by colonoscopy. Biopsy specimens were obtained from severely involved areas. If we could not detect abnormal findings, biopsy specimens were taken from normal appearing areas. Diagnosis of CMV enterocolitis and management of CMV reactivation have been reported previously.3 A univariate analysis using Fisher’s exact test was performed to compare differences in endoscopic findings between patients with and without CMV enterocolitis. A p value of <0.05 was considered significant.

CMV enterocolitis was diagnosed in eight patients; the remaining 25 patients served as controls. Symptoms in patients with and without CMV enterocolitis, respectively, were abdominal pain (n = 4 and n = 12), bloody stool (n = 1 and n = 6), and watery diarrhoea (n = 7 and n = 25). All of the patients with CMV enterocolitis tested positive for CMV antigenaemia with median levels of 84.5 per 50 000 cells (range 8–932) when they underwent colonoscopy. Eight controls tested positive for CMV antigenaemia. Median levels were 0 per 50 000 cells (range 0–17).

For endoscopic findings, patients with and without CMV enterocolitis, respectively, showed erosion (n = 5 and n = 5), ulceration (n = 1 and n = 1), oozing (n = 2 and n = 5), redness (n = 7 and n = 15), oedema (n = 1 and n = 6), disappearance of vascular network (n = 8 and n = 21), atrophic villi (n = 2 and n = 4), rough mucosa (n = 6 and n = 18), and exfoliation of mucosa (n = 3 and n = 10) (fig 1). Erosion was more frequent in patients with CMV enterocolitis than in controls (p = 0.036). Gastrointestinal GVHD was diagnosed in seven patients with CMV enterocolitis and in all of the controls. A total of 46 biopsy specimens were obtained from the eight patients with CMV enterocolitis (table 1). CMV inclusion bodies were found in 18 specimens, most of which were obtained from erosions (n = 13). CMV inclusion bodies were distributed widely from the terminal ileum to the rectum.

Table 1

 Endoscopic findings and localisation of cytomegalovirus (CMV) inclusion bodies in patients with CMV enterocolitis

Figure 1

 Endoscopic findings of cytomegalovirus (CMV) enterocolitis. (A) Erosion. (B) Erosion (after indigo carmine dye spray). (C) Ulceration. (D) Oozing.

The present study suggested endoscopic findings such as erosions were a useful marker for early diagnosis of CMV enterocolitis. Most of the CMV inclusion bodies were obtained from erosions. In contrast, only one had punched out ulcerations which had been considered characteristic of CMV enterocolitis.4–6 In case No 7, the only patient with punched out ulcerations, three specimens obtained from ulcerations were negative for CMV inclusion bodies while one specimen obtained from erosions showed inclusion bodies. CMV infected vascular endothelium is considered to narrow the vessels and to induce local ischaemia and eventually erosions and ulcerations.7 The small size of the erosions compared with that of ulcerations may enable biopsy of the whole lesion and the high yield of CMV infected cells. Biopsy of erosions can be reasonably effective for early diagnosis of CMV enterocolitis.

Colonoscopies in transplant recipients are often restricted to the rectum and sigmoid colon to minimise colonoscopic invasion. If we had examined only the rectum and sigmoid colon, CMV enterocolitis in six of the eight patients would have been missed. Total colonoscopy is necessary to make an early diagnosis of CMV enterocolitis. The present study suggests the usefulness and necessity of colonoscopy for an early diagnosis of CMV enterocolitis after allo-SCT.

References

Footnotes

  • Conflict of interest: None declared.