The Occurrence of Cytomegalovirus Gastrointestinal Disease in Patients with Gastrointestinal Graft-Versus-Host Disease and Its Related Clinical Outcomes in the Era of Preemptive Therapy for Cytomegalovirus Infection

Abstract 492

Diagnosis and therapeutic decision of gastrointestinal (GI)-graft-versus-host disease (GVHD) mainly rely upon clinical symptoms and exclusion of other causes of GI dysfunction. Particularly, the presence of CMV-GI-disease should be examined with immunostaining of biopsy. However, there have been few reports about the occurrence of CMV-GI-disease in patients with GI-GVHD in the era of preemptive therapy for CMV infection with real-time quantitative polymerase chain reaction (RQ-PCR) and its related outcomes.

Between January 2008 to December 2010, 392 adult patients with AML, ALL, MDS, and CML underwent allogeneic stem cell transplantation (SCT) and all had serial monitoring for CMV DNA based on RQ-PCR followed by initiation of preemptive therapy. Ninety-seven percent of recipients and 84% of donors were CMV-seropositive before SCT. Clinical GI-GVHD based on symptoms, stool studies, and endoscopic findings developed in 114 patients (29%) and they were treated by high-dose systemic steroid. Of these patients, 103 patients performed endoscopic biopsies not only at the time of onset of clinical symptoms but also during the clinical course of GI-GVHD in the absence of response to steroid, and all samples from biopsies were immunohistochemically stained. The median age of the 103 patients was 40 years (range, 15–70). These patients consisted of AML (n=59), ALL (n=29), MDS (n=12), and CML (n=3) transplanted from sibling (n=45), unrelated (n=48), familial-mismatched donors (n=8), and umbilical cord donors (n=2). Conditioning regimens consisted of myeloablative (76%) and reduced-intensity regimens (24%). Anti-thymocyte globulin was given as a part of conditioning in 41 patients (40%). GVHD prophylaxis was based on administering calcineurin inhibitor with short-course methotrexate. GI-GVHD was developed as a manifestation of not only acute GVHD (aGVHD, 87%) but also overlap syndrome (13%). Isolated GI-GVHD was 30%. Eighty-nine patients (86%) initially presented with GI-GVHD, whereas 14 patients (14%) developed GI-GVHD after the treatment of GVHD of other organs.

Of the 103 patients, 60 (58%) had histologic findings for GI-GVHD (lower gut, n=41, 40%; lower gut + upper gut, n=10, 10%; upper gut, n=9, 8%) and 26 (25%) had positive immunohistochemical stain for CMV-GI-disease (lower gut, n=13, 13%; lower gut + upper gut, n=4, 4%; upper gut, n=9, 8%). The CMV-GI-disease developed at a median time of 56 days after SCT (range, 20–403). Multivariate analyses revealed that older age over 30 years (RR 4.42, 95%CI 1.10–17.84), HLA-mismatched donors (RR 3.41, 95%CI 1.21–9.60), and higher grade of GVHD (grade III+IV aGVHD or severe cGVHD, RR 3.24, 95% CI 1.22–8.60) were significant risk factors for CMV-GI-disease. Most patients (73%) developed CMV-GI-disease after the treatment of GI-GVHD or other organs of GVHD, whereas CMV-GI-disease developed concurrently with GI-GVHD in other 27%. After the anti-viral therapy with ganciclovir or foscarnet, 14 patients (54%) showed complete resolution of CMV-GI-disease, while 12 patients (46%) failed to show any response.

There were 4 patterns of development of pathologically proven GI-GVHD and CMV-GI-disease (Figure). GVHD-specific survival of patients with CMV-GI-disease was significantly lower than patients without CMV-GI-disease (36.2% vs. 79.3%, P = 0.002). Patients with both GI-GVHD and CMV-GI-disease (n=18) had significantly inferior GVHD-specific survival than other 3 patterns of development of GI-GVHD and CMV-GI-disease (24.5% vs. 78.2%, P = 0.001).

Our data demonstrate 25 percent of patients with clinical GI-GVHD developed CMV-GI-disease during the treatment or at the time of diagnosis of GI-GVHD even in the era of preemptive therapy for CMV disease based on RQ-PCR, especially in older patients, HAL-mismatched transplants, and higher grade of GVHD. Patients with CMV-GI-disease, especially who concurrently had pathologically proven GVHD, had inferior GVHD-specific survival. Thus, we reemphasize the need of endoscopic biopsy with immunohistochemical stain for CMV disease if ever possible at the diagnosis of GI-GVHD as well as in the absence of response to treatment of GVHD, especially in high risk groups, which can facilitate proper selection of treatment. Additionally, novel prophylactic measures such as immunotherapy and drug prophylaxis should be considered in this specific group of patients.

View largeDownload slide

View largeDownload slide

Close modal