Introduction
The current treatment guidelines for ulcerative colitis (UC) recommend topical or systemic corticosteroids when the disease is refractory to 5-aminosalicylic acid (5-ASA) preparations (1). Because budesonide undergoes extensive first-pass hepatic metabolism, thereby reducing systemic adverse effects, it has been emphasized as an induction option for mild-to-moderate UC (2,3). Cytomegalovirus (CMV) reactivation is a recognized contributor to steroid refractoriness, with reported rates of approximately 30-60% among patients treated with systemic corticosteroids (4,5).
In contrast, CMV disease during locally acting budesonide therapy has been rarely documented, with reports largely outside UC; for example, CMV esophagitis during topical/swallowed budesonide for eosinophilic esophagitis (6,7), and its clinical significance remains unclear. We herein report a case of UC in which CMV colitis occurred while the patient was receiving budesonide, highlighting the need to consider early CMV testing before escalation to systemic corticosteroids.
Case Report
The patient was a 69-year-old woman with a history of appendicitis, meningitis, and dyslipidemia. She smoked 20 cigarettes per day for 40 years and did not drink alcohol. The patient had no family history of inflammatory bowel disease.
In May 2024, she developed increased stool frequency and hematochezia, and was diagnosed with left-sided UC based on colonoscopy at a local hospital. Time-dependent-release mesalazine (4,000 mg/day) induced only transient remission; relapse prompted the addition of budesonide rectal foam (2 mg/day) and budesonide MMX (9 mg/day), which proved ineffective. She was referred to our department in March 2025.
Switching to mesalazine MMX (4,800 mg/day) provided little benefit, and she was hospitalized in April because of worsening symptoms.
At the referring hospital, budesonide MMX (9 mg/day) was introduced for relapsing left-sided colitis. After referral to our department, budesonide rectal foam (2 mg, twice daily) was added. Colonoscopy performed six days before admission showed severe left-sided colitis with a Mayo endoscopic subscore of 3 and longitudinal undermined (“double-tracked”) ulcers, and CMV pp65 antigenemia was positive; therefore, both budesonide preparations were discontinued because CMV colitis was strongly suspected, whereas mesalazine MMX (4,800 mg/day) was continued.
On admission, the patient's temperature was 36.7°C, her blood pressure was 97/72 mmHg, her pulse was 91/min, and left lower quadrant tenderness was noted. She had ≥10 bloody stools per day. At that time, laboratory tests showed a white blood cell count of 4,600 /μL and a C-reactive protein (CRP) level of 0.36 mg/dL, with anemia (hemoglobin 7.5 g/dL) and hypoalbuminemia (albumin 2.9 g/dL) (Table).
There was no history of primary immunodeficiency, malignancy, transplantation, or the use of systemic immunosuppressants or biologics other than locally acting budesonide. T-SPOT was negative, and β-D-glucan was low. A stool culture was negative for pathogenic Escherichia coli and both Clostridium difficile toxin and antigen; however, CMV pp65 antigenemia was positive (C7-HRP: 1 cell per 50,000 leukocytes).
Contrast-enhanced abdominopelvic computed tomography revealed edematous wall thickening from the rectum to the sigmoid colon, with enlarged regional lymph nodes. Colonoscopy revealed loss of the vascular pattern and longitudinal ulcers with easy bleeding in the same segment, yielding a Mayo endoscopic subscore of 3 (Fig. 1).
Deeply undermined “double-tracked” ulcers suggested CMV infection, and immunohistochemistry of biopsy specimens confirmed the presence of CMV-positive cells. Histological examination of targeted biopsy specimens from the rectosigmoid ulcer edge showed active inflammation and granulation tissue. Typical intranuclear inclusion bodies were not identified by routine hematoxylin and eosin (H&E) staining; however, CMV immunohistochemistry demonstrated a single CMV-positive cell with nuclear staining, confirming CMV colitis (Fig.
3A, B). Based on these findings, she was diagnosed with CMV reactivation complicating an exacerbation of UC during budesonide therapy. The overall clinical course of the patient is shown in Fig.
4.
Although she developed fever after admission, her stool frequency and CRP level did not worsen. Thus, intolerance to 5-ASA was therefore considered unlikely, and treatment was continued. Oral valganciclovir (900 mg, twice daily, total 1,800 mg/day) was initiated on hospital day (HD) 6, reducing her stool frequency to four times daily by day 8, with CMV pp65 antigenemia converting to negative.
Follow-up colonoscopy on HD 10 showed shallower ulcers with reduced friability in the rectosigmoid colon (the same segment as Fig. 1), consistent with mucosal improvement (Fig. 2).
Valganciclovir was discontinued after three weeks. Approximately six months later, she remained in remission with mesalazine monotherapy without relapse.
Discussion
In Japan, where UC is designated as an intractable disease by the Ministry of Health, Labour and Welfare, the patient population continues to rise and is now likely to exceed 300,000 when mild cases are included. Consequently, in addition to specialized centers, general practitioners who perform endoscopy increasingly manage UC, often using foundational agents such as 5-ASA and corticosteroids. The 2023 governmental guidelines specified oral budesonide MMX (enteric-coated extended-release) in addition to the previously available budesonide enema for mild-to-moderate UC (8).
Budesonide is an “antedrug” steroid that exerts efficacy locally at inflamed sites and is rapidly metabolized after absorption, thereby minimizing systemic adverse effects (9). Nevertheless, in our case, CMV colitis became clinically apparent during budesonide therapy. However, it remains uncertain whether budesonide directly induces viral reactivation.
Several interpretations are possible: CMV infection may have already been present before the initiation of budesonide, underlying disease activity might have promoted CMV reactivation, or partial systemic absorption of budesonide could have contributed to transient immunosuppression. The sustained remission achieved with mesalazine monotherapy after antiviral therapy suggests that CMV infection, rather than budesonide refractoriness, was the main driver of the clinical worsening. Because this was an elderly-onset case of UC, age-related immunosenescence and comorbidities may have modestly increased the susceptibility to CMV reactivation, even under treatment with a locally acting steroid.
In this context, an accurate diagnosis of CMV colitis is crucial. Our patient's pp65 antigenemia titer was low (C7-HRP: 1 cell per 50,000 leukocytes), yet the diagnosis was supported by the combination of characteristic colonoscopic morphology, biopsy-proven CMV-positive cells, and a poor response to budesonide. Consistent with prior studies, pp65 antigenemia and quantitative PCR have imperfect sensitivity and specificity and use heterogeneous cutoff values; therefore, serologic assays should not be interpreted in isolation when assessing CMV in UC (4,10,11).
In our case, the rapid clinical response to valganciclovir with antigenemia converting to negative by day 8 further supports the pathogenic relevance of CMV despite the low initial titer.
Two non-mutually exclusive mechanisms may explain why CMV colitis became clinically apparent when the patient was receiving extended-release budesonide. Although budesonide MMX is engineered for colonic delivery with extensive first-pass hepatic metabolism, some systemic absorption cannot be excluded (2,3), and corticosteroid exposure has been associated with CMV reactivation and steroid-refractory disease in UC (4,11). Second, a local mucosal pathway may operate in parallel; high topical glucocorticoid concentrations at inflamed sites can modulate mucosal immune responses and could, under certain conditions.
In our case, biopsy-proven CMV-positive cells with low pp65 antigenemia suggested a process largely localized to the colonic mucosa, although the relative contribution of systemic versus local steroid effects cannot be determined from a single case (10,11). Analogously, CMV esophagitis has been described during topical or swallowed budesonide treatment for eosinophilic esophagitis, where local steroid stasis at a stenotic lesion was implicated, further supporting a potential local mucosal mechanism under high topical exposure (6,7).
There is no universal consensus on the initiation of antiviral therapy for CMV reactivation in patients with UC. Proinflammatory cytokines can promote the differentiation of latently infected cells, and both immunosuppressive therapy (including steroids) and underlying mucosal inflammation may influence reactivation (12). Low-grade mucosal CMV signals may occasionally abate with control of inflammation alone, but current evidence generally supports initiating antiviral therapy when tissue-invasive CMV is documented in steroid-refractory or steroid-dependent disease or when systemic CMV activity accompanies typical endoscopic lesions, in which antivirals have been associated with improved outcomes and may prevent unnecessary escalation of immunosuppression (12).
In hospitalized patients with severe UC, steroid nonresponse remains common (13), and CMV reactivation rates have been reported to be as high as ～50% among steroid-refractory cases (8).
Our case illustrates that these considerations apply to patients receiving budesonide MMX. Budesonide MMX and 5-ASA failed to induce remission, and colonoscopy shortly before admission revealed deeply undermined (“double-tracked”) ulcers with CMV-positive biopsies and pp65 antigenemia. Ordinarily, in budesonide MMX-treated patients who show no response or early worsening, it may be prudent to evaluate CMV first, using endoscopy with targeted biopsies for H&E and CMV immunohistochemistry, supported by selective virologic assays, rather than automatically escalating to systemic corticosteroids, because active CMV may be driving symptoms and targeted antiviral therapy can rapidly improve the course and avert progression to steroid-refractory disease.
In our patient, targeted valganciclovir allowed for rapid improvement and obviated unnecessary treatment escalation to biologics or other advanced therapies.
Based on our experience and prior reports, a pragmatic checkpoint for CMV assessment in UC during budesonide therapy could include: (i) Who (patients with persistent symptoms or early worsening despite budesonide, particularly when endoscopy shows deep undermined or longitudinal ulcers and escalation to systemic steroids or advanced agents is being considered); (ii) When [at the point of clinical non-response or deterioration (e.g., failure to improve after a short induction period or frank worsening), before the initiation of systemic corticosteroids]; and (iii) How to perform colonoscopy with targeted biopsies (including ulcer edges and bases) for H&E and CMV immunohistochemistry, supported by pp65 antigenemia and/or quantitative polymerase chain reaction, while recognizing their imperfect performance characteristics (4,9-11). This structured approach aims to minimize unnecessary escalation while identifying patients who may benefit from timely antiviral therapy and highlights that locally acting budesonide MMX does not eliminate the risk of CMV colitis. Reactivation can still occur during therapy and should not be dismissed solely because a locally acting formulation is being used.
Written informed consent for publication of the clinical details and images was obtained from the patient.
The authors state that they have no Conflict of Interest (COI).