CMV ileal ulcers cause massive bleeding in kidney transplant recipient

A 26-year-old kidney transplant recipient experienced massive gastrointestinal bleeding caused by cytomegalovirus (CMV) ulcers in the distal ileum, according to a 2026 case report. He passed fresh and altered blood as many as six times a day and was admitted with a hemoglobin concentration of 5.1 g/dL, requiring transfusion support. Urgent upper endoscopy, computed tomography arteriography and colonoscopy did not identify a bleeding lesion. Capsule endoscopy later showed blood and clots throughout much of the ileum and colon, supporting a small-bowel source. Double-balloon enteroscopy, not capsule imaging, ultimately revealed several ileal ulcers whose biopsies established tissue-invasive CMV disease.

The patient had received a renal transplant in 2021 for C1Q nephropathy with focal segmental glomerulosclerosis and was receiving mycophenolate mofetil, tacrolimus and prednisolone. He also had thalassemia minor, with a baseline hemoglobin level of approximately 10–11 g/dL. At admission, he reported general malaise consistent with anemia but denied abdominal pain, vomiting and fever. His clinical history included no suspicious food exposure, relevant travel or alcohol use, and no use of over-the-counter drugs or herbs. The case demonstrates how a sparse, small-bowel presentation of CMV can escape detection during an initial evaluation of major gastrointestinal bleeding in an immunosuppressed transplant recipient.

Initial examination found altered blood on digital rectal examination, while proctoscopy showed bleeding from above the anal canal without a visible rectal or anal lesion. Urgent oesophagogastroduodenoscopy on the first day of admission showed no active bleeding and no identifiable lesion. Computed tomography arteriography performed the same day likewise found no contrast extravasation or gastrointestinal source. The following day, colonoscopy reached 20 cm proximal to the ileocaecal valve and revealed pools of blood-stained fluid with clots along the entire colon. Only trace blood-stained fluid was seen beyond the ileocaecal valve, and no mucosal lesion was identified.

Because the standard endoscopic examinations had not located the source, clinicians suspected small-bowel bleeding. Capsule endoscopy on day five showed that most of the ileum and the whole colon contained altered blood and a large quantity of clots, further supporting that suspicion. The patient continued to pass blood one to three times per day. A total of 10 units of packed red blood cells were administered during his first five days in hospital, and his hemoglobin was maintained at approximately 7–8 g/dL. The capsule findings indicated the distribution of blood but did not identify the ulcerated lesions, which were detected only during the later double-balloon enteroscopy.

The renal team advised discontinuing mycophenolate mofetil on day eight while continuing tacrolimus and prednisolone. The drug had been associated with small-bowel bleeding in transplant recipients, and it suppresses lymphocyte proliferation, which may increase the risk of CMV reactivation. The patient’s tacrolimus concentration was within the therapeutic range. Because drug-induced small-bowel injury remained a possibility and double-balloon equipment was unavailable, clinicians also started off-label misoprostol at 200 micrograms four times daily. This prostaglandin analogue is approved for non-steroidal anti-inflammatory drug-induced gastric ulcers and has evidence in aspirin- or NSAID-associated small-bowel ulcers; the authors considered that it could support mucosal defence and cellular repair.

Bleeding ceased by days nine and 10, before valganciclovir was started, and the patient’s hemoglobin rose to about 9 g/dL. The clinical improvement coincided with withdrawal of mycophenolate mofetil, but the single-patient report cannot establish that this change alone stopped the bleeding. The authors suggest that reducing immunosuppression may have permitted partial restoration of the T-lymphocyte-mediated control of CMV replication. Later negative CMV DNA results before antiviral treatment provide a possible supporting observation, although blood viral levels do not necessarily correspond to the activity of tissue-invasive disease. The bleeding stopped and the anemia improved, but the individual contribution of each intervention remained uncertain.

Double-balloon enteroscopy through the anal route was performed on day 12, after the overt bleeding had subsided. Enteroscopy reached 80 cm proximal to the ileocaecal valve and identified several ulcers in the distal ileum, the largest measuring 1.5 cm in diameter. The ulcers had regular margins, sloping edges, clean bases and mild surrounding erythema, with blood clots mildly adherent to their surfaces. No active bleeding was present during the examination. Biopsies showed small-intestinal mucosa with marked inflammation, CMV viral inclusions and a positive CMV immunostain. The investigators diagnosed CMV small-bowel ulcer, classified as tissue-invasive CMV disease, responsible for the substantial gastrointestinal bleeding.

Blood testing presented another diagnostic limitation. A CMV pp65 antigen assay on day eight was negative, whereas CMV DNA PCR on the same day returned an elevated result of 1.90 × 10^3 IU/mL. The pp65 assay detects CMV antigen inside leukocytes and can be falsely negative in immunosuppressed patients. PCR detected circulating viral DNA but could not by itself locate CMV disease or prove that the ileal lesions were responsible for the bleeding. Establishing the diagnosis therefore depended on direct visualization and tissue obtained during enteroscopy. The subsequent disappearance of viraemia also did not necessarily mean that the ulcers had healed, an important distinction in a tissue-invasive infection.

CMV is human herpesvirus 5, a virus distributed worldwide and usually acquired in childhood or early adulthood. It can replicate in the nuclei of epithelial cells, endothelial cells, fibroblasts and leukocytes, and after the initial phase it can remain latent in myeloid cells in bone marrow. Reactivation may occur later, particularly when cell-mediated immunity is impaired, as it was in this transplant recipient. The presence of active viral replication is termed CMV infection, whereas symptomatic infection is termed CMV disease. The gastrointestinal tract accounts for about 30% of tissue-invasive CMV disease and can be involved from the mouth to the anus. The colon and rectum are the most frequently affected sites, followed by the stomach and oesophagus, making small-bowel involvement comparatively rare.

Reported CMV small-bowel disease has several morphological patterns. Diffuse swelling with erosions is commonly described and may be visible on computed tomography when marked mucosal swelling is present. Discrete erosions and ulcers can also occur in multiples over a long segment of the gastrointestinal tract. The few small, separate ulcers found in this patient represent a less commonly reported appearance and are difficult to detect. An ulcer without substantial surrounding oedema may not be visible on cross-sectional imaging, while small, sparsely distributed lesions can also be missed during endoscopy. Untreated ulcers may progress, with bleeding described as the most common complication, although some cases present with perforation requiring surgery. The delay in this case—from suspicion of small-bowel bleeding on day five to enteroscopy on day 12—reflected the technical demands and limited local availability of double-balloon equipment and expertise.

After discharge on day 18, the patient was reviewed four days later. Given the tissue-invasive CMV ulcers and his immunocompromised status, clinicians began oral valganciclovir 450 mg daily at a renal-adjusted dose for three weeks. A CMV DNA PCR sample collected before the first antiviral dose was already negative, as were subsequent tests at weeks one and three. The authors suggest that reduced immunosuppression may have contributed to the clearance of viraemia, while noting that this finding did not establish that the ulcers had healed. Weekly follow-up for two additional weeks recorded no further bleeding and stable hemoglobin. Colonoscopy one month after the index enteroscopy showed erythema and scarring around the ileocaecal valve, together with patchy terminal-ileal erythema considered consistent with the previous ulcers. This case documents resolution of overt bleeding and improvement in anemia, but a single-patient report cannot establish the frequency of CMV ileal bleeding or the comparative effectiveness of withdrawal of mycophenolate, misoprostol and antiviral treatment.

Subject of Research: A case report of tissue-invasive CMV distal-ileal ulcers causing massive gastrointestinal bleeding in a renal transplant recipient.

Article Title: A rare cause of massive gastrointestinal bleeding in a renal transplant recipient—CMV small-bowel ulcer: a case report

Article References: Lam, T., & Yuen, K. Y. (2026). A rare cause of massive gastrointestinal bleeding in a renal transplant recipient—CMV small-bowel ulcer: a case report. Frontiers in Medicine, 13, Article 1959416. https://doi.org/10.3389/fmed.2026.1959416

Image Credits: AI Generated

DOI: 10.3389/fmed.2026.1959416

Keywords: cytomegalovirus, small-bowel ulcers, gastrointestinal bleeding, kidney transplantation, immunosuppression, double-balloon enteroscopy, ileal biopsy, valganciclovir, CMV reactivation, case report, rare, cause