This is the first study to evaluate the relationship between HCMV-DNA copy numbers in colonic tissues and other HCMV tests or endoscopic scores. We found a significant correlation between HCMV-DNA copy number in colonic tissues and both the antigenemia assay and histopathological tests by IHC in UC patients with concomitant HCMV colitis. Regarding endoscopic scores, HCMV-DNA copy number by mucosal PCR correlated with both the total UCEIS score and the bleeding score. Moreover, a positive correlation between the change in the total UCEIS score and HCMV-DNA copy number for UC patients taking multiple mucosal PCR tests.

Generally, HCMV is contracted during childhood and can persist as a lifelong latent infection, reactivating under inflammatory conditions in an immunosuppressed host [1, 2]. HCMV infection is involved in active UC patients who are refractory to steroids and immunosuppressive drugs [3, 5, 18]. Previous reports showed that steroids are predisposing factors for HCMV reactivation by suppressing anti-HCMV T-cell specific function and immunoglobulin production via B-cells, and by directly activating viral replication [19,20,21]. In this study, patients treated with PSL had a higher HCMV-DNA copy number than those treated without PSL. In patients with acute steroid-resistant UC, HCMV infection should be excluded before increasing immunosuppressive drugs [22]. However, the clinical features of HCMV enterocolitis are sometimes indistinguishable from UC relapse [23,24,25].

Histological examination, including H&E staining and IHC, on biopsy from colonic tissues is the gold standard for diagnosing HCMV enterocolitis, which yields high specificity but variable sensitivity [1, 6, 7]. IHC has 22.1–30.0% higher sensitivity than H&E staining for diagnosing HCMV infection in IBD patients [26,27,28]. In the present study, the positive rate of HCMV antigen by IHC correlated with the HCMV-DNA copy number.

Another standard method for detecting HCMV infection is the HCMV antigenemia assay. Previous reports demonstrated that the antigenemia assay has a high specificity for detecting HCMV and a significant association with the subsequent colectomy rate in UC patients with concomitant HCMV enterocolitis [8,9,10,11]. However, the sensitivity for diagnosing HCMV gastrointestinal diseases is only 47.0–67.3%, suggesting that the antigenemia assay alone is considered insufficient for early detection of HCMV [8,9,10,11].

Yoshino et al. reported that a mucosal quantitative real-time PCR assay for detecting HCMV-DNA in the gastrointestinal tract was useful for the early and accurate diagnosis of HCMV infection in active UC patients refractory to immunosuppressive therapies [12]. Roblin et al. reported that UC patients with an HCMV-DNA load higher than 250 copies/mg in colonic tissue required early antiviral treatment [29]. However, mucosal PCR assays cannot be available in all medical institutions, and most physicians should decide to start antiviral treatment based on the results of antigenemia assays or histopathological tests in clinical practice. In the present study, we found a positive correlation between HCMV-DNA copy number and both HCMV antigenemia assay and IHC positivity. ROC analysis showed that 1300 (AUC = 0.80, 95% CI = 0.64–0.90) and 1,650 (AUC = 0.81, 95% CI = 0.65–0.98) copies/µg of HCMV-DNA were the best diagnostic values to detect HCMV antigenemia-positive cells and IHC positivity, respectively. These data suggest that positivity for HCMV antigenemia and IHC signify the increase of HCMV-DNA in the inflamed colon to decide the intervention of antiviral treatment for HCMV.

Conversely, among UC patients who were negative for both the antigenemia and histopathological tests, the HCMV-DNA copy numbers were significantly lower than those in the other groups. In this group, the median HCMV-DNA copy number was 300 copies/μg (interquartile range: 39–810 copies/μg); however, there were no significant differences in HCMV-DNA copy numbers and the clinical outcome regardless of the antiviral treatment. These data suggest that a group of patients with a low mucosal HCMV-DNA copy number (less than 800 copies/μg) during treatment is more likely to be controlled with immune regulating therapy alone. However, given that the inflammatory control was poor during treatment, even in patients with a low mucosal HCM-DNA copy number at diagnosis, we must consider that exacerbation of HCMV infection can modify the disease state.

Regarding endoscopic findings, Suzuki et al. reported that irregular punched-out or longitudinal ulcers were specific for HCMV infection in UC patients [24]. On the other hand, the endoscopic findings of HCMV colitis are variable and sometimes do not reveal specific features [12, 30]. The present study also showed that the endoscopic scores varied even with the same amount of HCMV-DNA copy numbers. Taken together, previous and current data suggest that the heterogeneity of the immune response to HCMV and mucosal cytokine pattern in each UC patient could contribute to endoscopic features of UC with concomitant HCMV infection [5, 31,32,33]. To date, there have been no reports on the relationship between HCMV enterocolitis and endoscopic scores for MES or UCEIS in IBD patients. In this study, the total UCEIS score and the bleeding score were correlated with the HCMV-DNA copy numbers. Moreover, when we focused on patients who underwent multiple HCMV mucosal PCR tests, we found a significant correlation between changes in HCMV-DNA and those in the total UCEIS. At the beginning of HCMV reactivation, HCMV can infect the vascular endothelium and cause ischemic damage to the mucosa, which results in gastrointestinal bleeding [34]. Previous reports have shown that gastrointestinal bleeding is the most frequent symptom of HCMV enterocolitis [34, 35]. Thus, in UC patients refractory to immune suppressive therapy who had a higher total UCEIS score, severe bleeding score, and increasing total UCEIS score during the clinical course, we should consider the involvement of HCMV reactivation in the colon. Furthermore, HCMV-DNA copy number tended to increase in cases who led to colectomy. These results suggest that early initiation of antiviral therapy may avoid the risk of colectomy in cases with elevated UCEIS scores.

There are several limitations to this study. First, because of the retrospective study design, not all patients underwent antigenemia assays and histopathological tests with IHC. Additionally, not all HCMV antigenemia assays were performed on the same date as the biopsy for mucosal PCR. Since the difference in dates between HCMV antigenemia and mucosal PCR tests was at most 7 days, of which 71% were within 3 days, the initiation of therapeutic agents (i.e., steroids, biologics, and antiviral agents) might have affected the value of antigenemia in cases where the antigenemia was measured after mucosal PCR was performed. Second, whether to measure HCMV-DNA was made by each attending physician, which could be at risk for selection bias.

In conclusion, our data showed that the antigenemia assay and histopathological test with IHC could help estimate the HCMV-DNA copy number in colonic tissues. Furthermore, the changes in UCEIS score in each individual may be helpful for decision-making in UC patients with concomitant HCMV infection. Additional study is required to establish an accurate diagnostic scoring system for the early diagnosis of HCMV infection in UC patients with a combination of HCMV tests and endoscopic score.