Is Follow-Up Endoscopy Necessary in Upper Gastrointestinal Cytomegalovirus Disease?

Abstract

Gastrointestinal (GI) cytomegalovirus (CMV) disease is a major cause of morbidity and mortality in immunocompromised patients. Diagnosis of GI CMV disease mostly relies on endoscopy examination and histopathologic findings. There are limited data on the need for follow-up endoscopy with histopathologic examination in patients with upper gastrointestinal (UGI) CMV disease. All adult patients with confirmed and probable UGI CMV disease at a tertiary hospital over a 16-year period whose follow-up endoscopy was available were enrolled. The patients were classified as endoscopic responders if they showed complete or partial improvement on follow-up endoscopy, and as endoscopic nonresponders if there was no improvement or worsening. CMV tissue clearance was defined as absence of any visible CMV inclusion bodies, negative CMV immunohistochemistry and negative CMV polymerase chain reaction in follow-up biopsy tissues. During the study period, 77 patients with UGI CMV disease were analyzed. The median time to follow-up endoscopy was 19 days (interquartile range, 14-27). Of these 77 patients, 52 (68%) were classified as responders, and the remaining 25 (32%) as nonresponders. GI bleeding was more common in the nonresponders than the responders (36% vs 12%, respectively; P = 0.02). There was no significant difference in CMV tissue clearance between the responders and nonresponders (56% vs 69%, respectively; P = 0.38), median durations of treatment (20 days vs 21 days, respectively; P = 0.48), and relapse rates (10% vs 8%, respectively; P > 0.99). Multivariate analysis showed that the only independent predictive factor for relapse of CMV antigenemia or CMV GI disease was multiorgan CMV disease (odds ratio = 12.4, 95% confidence interval 1.6-97.9; P = 0.02). Endoscopic responses were obtained in about two-thirds of patients with UGI CMV disease 2 or 3 weeks after antiviral therapy. However, these follow-up endoscopic findings neither reflected CMV tissue clearance nor predicted disease relapse. These findings suggest that the routine follow-up endoscopy may not be warranted in patients with UGI CMV disease.

FIGURE 1

Flow chart of patients. CMV = cytomegalovirus, UGI = upper gastrointestinal. ^aConfirmed UGI CMV disease was defined as symptoms or signs of gastrointestinal disease with positive pathologic findings of CMV infection from hematoxylin and eosin (H&E) staining and immunochemical staining of endoscopic biopsies. ^bProbable disease was defined as a positive CMV polymerase chain reaction from a biopsy specimen showing clinical improvement with antiviral therapy.

FIGURE 2

Kaplan–Meier curves of cumulative percentages of viral clearance from blood (A) and tissue (B) in endoscopic responders and nonresponders. The time to blood (A) and tissue (B) clearance did not differ between the 2 groups.