For patients with immune checkpoint inhibitor–induced colitis, Epstein-Barr virus (EBV) infection may increase risks of steroid-refractory disease and ulcers that contribute to colonic perforation, according to investigators.
Pending further research, routine monitoring of EBV status may be needed for patients undergoing checkpoint inhibitor therapy, reported lead author Matthew R. Pugh, FRCPath, of University Hospital of Wales, Cardiff, and colleagues.
“Few studies have investigated the role of viruses in the pathogenesis of immune-related colitis,” the investigators wrote. Their report is in Clinical Gastroenterology and Hepatology. While cytomegalovirus has been linked with worse disease, no studies to date have evaluated the role of EBV, they noted, despite theoretical concerns.
“A spectrum of EBV-positive lymphoproliferations shows a predilection for the GI tract, ranging from indolent lesions to aggressive lymphomas,” the investigators wrote. “One such proliferation, EBV-positive mucocutaneous ulcer (EBVMCU), is an indolent, ulcerating process associated with immunosuppression,” they added, referring to studies involving patients with inflammatory bowel disease.
To determine if EBV could be playing a similar role in cancer immunotherapy, the investigators retrospectively analyzed colon tissue samples from 16 patients who developed colitis after undergoing immune checkpoint inhibitor therapy between 2010 and 2018. Thirteen patients received an anti-CTLA-4 agent, three were treated with a PD-1 inhibitor, and four received both types of therapy. Most patients had advanced-stage melanoma (n = 14), while the remaining two patients had prostate and renal carcinoma, respectively. Ten samples were biopsies, whereas four specimens were collected from surgical repair of colon perforation.
EBV status was determined by chromogenic in situ hybridization for EBV-encoded small RNA, with positive samples further characterized by immunohistochemistry for CD3, CD15, CD20, CD30, CD138, MUM1, and PAX5. In addition, all samples were immunostained for cytomegalovirus, and PCR was used to assess B cell and T cell clonality.
The median time from induction of therapy to colitis onset was approximately 1 month (32.5 days), with symptoms typically lasting 3 weeks (22.5 days). Macroscopically, 10 patients had ulceration, and 6 displayed signs of hemorrhage.
EBVMCUs were found in four patients, of whom three had received anti-CTLA-4 therapy, one had received both anti-CTLA-4 and anti-PD-1 therapy, and all had undergone colonic resection. One case also tested positive for cytomegalovirus.
Immunostaining showed that EBVMCUs had underlying B cell and linear plasma cell infiltrates, with “a rim of small T lymphocytes at the base.” EBV-encoded small RNA expression was found in both plasma cells and small B cells.
The presence of EBVMCUs was significantly associated with more severe colitis.
All four EBV-positive patients had steroid-refractory colitis, compared with only two (12.5%) of the EBV-negative patients (P = .008), a difference that was echoed by the rate of colonic resection (100% vs. 12.5%; P .008). Furthermore, colon perforation occurred in all EBV-positive patients, versus none of the EBV-negative patients (P = .001).
For three EBV-positive patients, preresection biopsy samples were available, allowing for temporal analysis of EBV-encoded small RNA. Earlier samples had reduced or absent EBV-positive lymphoid cells, which offered some etiologic insight.
“The apparent absence or paucity of EBV-positive lymphoid cells in biopsies taken before resection suggests that EBVMCU is arising within preexisting immune-mediated inflammation rather than EBV driving the initial inflammatory insult,” the investigators wrote.
They suggested that EBVMCUs “likely contribute directly to colonic perforation,” since lesions are characterized by a form of localized tissue destruction that has been previously associated with colonic perforation in Crohn’s disease and intestinal perforation in rheumatoid arthritis.
Still, mechanisms of action remain unknown. “It is unclear why EBVMCUs should arise in the context of immune checkpoint regulator therapy, which, in contrast to conventional immunosuppressants, results in immune activation,” the investigators wrote. “It is possible that these patients may harbor residual immunosuppression resulting from their disease burden, advanced age, and prior immunosuppression.”
While more work is needed, Dr. Pugh and colleagues suggested that EBV testing may be valuable for some patients.
“The findings support the need for further studies investigating the role of EBV monitoring in immune checkpoint regulator therapy, which is not currently part of routine protocols.”
The study was funded by All Wales Lymphoma Panel. The investigators disclosed no conflicts of interest.
SOURCE: Pugh et al. Clin Gastroenterol Hepatol. 2019 Oct 11. doi: 10.1016/j.cgh.2019.09.031.