https://orcid.org/0000-0002-5148-6130
To the editors,
A 65-year-old male with chronic myeloid leukemia (CML) taking imatinib developed severe recurrent diarrhea, abdominal pain, and hematochezia. Medical history was notable for severe chronic obstructive pulmonary disease, congestive heart failure, and cerebrovascular accident. The patient’s CML was initially treated with dasatinib and later replaced with imatinib because of recurrent pleural effusion despite dose reduction. While this regimen attained complete cytogenetic and molecular response, the patient presented with gastrointestinal symptoms suspicious for colitis after 5 months of treatment. Bloodwork was notable for anemia, thrombocytopenia, moderate elevation of inflammatory markers, and hypoalbuminemia. Contrasted imaging of the abdomen showed a dilated right colon and transverse colon wall thickening (Supplementary Figs. 1, 2). Stool testing for infectious etiology was negative. A colonoscopy demonstrated Mayo grade 2 pancolitis confirmed on pathology that was most representative of ulcerative colitis.
Despite withheld chemotherapy and a trial of mesalamine, vedolizumab, and high-dose steroids, colitis symptoms persisted with a need for biweekly blood transfusions (Figs. 1, 2). The patient’s disease course was further complicated with Clostridium difficile and cytomegalovirus colitis, for which he was treated appropriately. With the dearth of treatment options given concurrent comorbidities and failure to achieve clinical or endoscopic remission, he was treated with compassionate fecal microbiota transplantation to which he responded clinically and endoscopically for 2 months. Unfortunately, thereafter, the patient clinically deteriorated with significant refractory anemia, and as a last resort he underwent a total colectomy (Supplementary Fig. 3). Pathology of the resected specimen confirmed ulcerative colitis. He did well from a gastrointestinal standpoint, and CML remained in remission on resumed imatinib therapy. Unfortunately, 12 months later the patient succumbed to respiratory complications with underlying chronic obstructive pulmonary disease.
Colonoscopy findings at different treatment time points. A, Erythema, edema, and pseudopolyps after steroid treatment and C. difficile and cytomegalovirus complications. B, Erythema, edema, superficial ulceration, loss of vasculature, and pseudopolyps after reduction of imatinib dosage and 5 doses of vedolizumab. C, Persistent mucosal inflammation after 7 doses of vedolizumab.
Pathological features of the colon biopsy. A, Neutrophil infiltration of the crypt, and lumen abscess consistent with active colitis (arrows). B, Lymph-plasma cell infiltration and crypt distortion (branching) consistent with chronic colitis (arrows).
Severe gastrointestinal inflammation can occur in cancer patients as a coexisting condition or cancer therapy–related toxicity, which can negatively impact a patient’s cancer journey. Data on the incidence of inflammatory bowel disease and guidelines on its management in this patient population on immunosuppressive therapy are limited, leaving most physicians in a clinical dilemma.1–3 With the limited immunosuppressive inflammatory bowel disease medications available to this group, surgery, albeit aggressive, remains a viable option. Finally, while the role of dasatinib inducing colitis has been explored,4, 5 the evidence of the role of imatinib remains unknown. Our case report may warrant further investigation if similar reports accrue in the future.
SUPPLEMENTARY DATA
Supplementary data are available at Inflammatory Bowel Diseases online.
Supplementary Figure 1. Dilated ascending colon.
Supplementary Figure 2. Wall thickening along the transverse colon as depicted by arrows and dilated ascending colon.
Supplemental Figure 3. Timeline of patient’s ulcerative colitis treatment course with fecal microbiota transplantation as reference point.
Supported by: The study was not supported by any funding.
Conflict of interest: Yinghong Wang serves as a consultant for Tillotts Pharma. The other authors declare no conflict of interest related to the study findings.
REFERENCES
Author notes
These authors contributed equally.
