Office-based flexible sigmoidoscopy allows rapid assessment and management of suspected immune checkpoint inhibitor-related colitis

Introduction

Immune checkpoint inhibitors (ICIs) have revolutionized cancer therapy and have improved survival for patients with various advanced or metastatic cancers [1]. ICIs targeting PD-1, PD-L1 and CTLA-4 inhibit checkpoint protein interactions and therefore enhance anti-tumor immunity. However, off-target inflammation and autoimmunity can cause toxicity in virtually any organ system [1,2]. These adverse events are collectively referred to as immune-related adverse effects (irAEs), while immunotherapy-related colitis (irColitis) is among the most common and the most frequent irAE leading to treatment discontinuation [3].

Workup of suspected irColitis should be rapid, as early identification and prompt treatment can guide the initiation of appropriate therapy, avoid detrimental colitis outcomes and allow thoughtful reintroduction of ICI when appropriate [2,4]. Together with stool testing for infection and inflammation, early endoscopy and tissue biopsy are essential for correct diagnosis, risk stratification and therapy guidance [2,3]. Current ASCO guidelines for the management of suspected irColitis recommend considering endoscopy or colonoscopy for patients with grade 2 or greater diarrhea, based on common terminology criteria for adverse events, as more severe mucosal inflammation has been linked to steroid refractoriness [5]. As >95% of irColitis cases are thought to involve the left colon, it has been suggested that flexible sigmoidoscopy with biopsies is sufficient for an initial evaluation, and may be more feasible for timely workup [6,7]. Despite multiple guidelines recommending early endoscopic evaluation for suspected irColitis, the real-world feasibility and impact on patient care remain unclear. The aim of this study was to assess the feasibility and impact on patient care of a rapid-access program for office-based sigmoidoscopic evaluations in patients with suspected irColitis.

Patients and methods

Results

Patient population

A total of 94 patients underwent office-based sigmoidoscopy for the initial evaluation of suspected irColitis. The baseline characteristics for patients who underwent initial evaluation are summarized in Table 1. Fifty-five (59%) patients were male, and the median age (IQR) was 67 (58-74) years. Melanoma was the most common cancer type (29%), followed by genitourinary (28%), lung (26%), and gastrointestinal (9%) malignancies. Anti-PD-(L)1 monotherapy was the most common therapy (56%), followed by combination CTLA-4/PD-1 therapy (41%). The median (IQR) time from GI referral to evaluation, including flexible sigmoidoscopy, was 8 (range 5-15) days.

Table 1 Characteristics of patients undergoing office-based flexible sigmoidoscopy for initial evaluation of suspected irColitis (n=94)

Alternative diagnoses

Alternative causes of GI symptoms diagnosed among patients with a negative workup for irColitis (n=28) can be found in Table 2. Common among these were concomitant cancer medications (29%), intestinal infections/diverticular disease (18%), cancer infiltration in the GI tract (14%) or newly diagnosed exocrine pancreatic insufficiency attributed to ICI (14%).

Table 2 Etiologies of GI symptoms among patients with initial exam negative for irColitis (n=28)

Discussion

Our study highlights the importance of early referral and rapid endoscopic evaluation of suspected irColitis to confirm the diagnosis and guide appropriate management. We report a median time of 8 days from GI referral to evaluation, including flexible sigmoidoscopy, suggesting that rapid evaluation is feasible via a dedicated program. Furthermore, we observed sigmoidoscopy to have excellent accuracy in determining the cause of lower GI symptoms for patients on ICI, with a clear impact on patient management.

The use of flexible sigmoidoscopy is beneficial, as it can be performed rapidly, without the need for sedation or a full bowel prep, and it is sufficient to diagnose and assess severity in the vast majority of irColitis cases. A prior systematic review found that 98% of patients with irColitis who underwent endoscopic evaluation had disease detected in the left colon [6]. Indeed, no cases initially diagnosed as “not irColitis” in our series were subsequently found to have a more proximal inflammatory toxicity that was missed on the sigmoidoscopic exam. Our study also confirms the importance of endoscopic evaluation for a correct diagnosis, as 30% of patients did not have irColitis detected on sigmoidoscopy: most of these avoided steroids and many were able to resume ICI.

Finally, prior studies have suggested that early endoscopic evaluation is essential for appropriate disease management. One study found that patients who underwent endoscopic evaluation after 30 days, compared to within 30 days from symptom onset, required a longer duration of steroids, had more recurrent symptoms and received later biologic add-on therapy, which has been associated with poorer outcomes [3,10,11]. Similarly, in our series 89% of patients with confirmed irColitis underwent changes to their colitis treatment regimen following endoscopic evaluation: 39% of the patients initiated medications for colitis treatment and 50% had changes in their treatment regimen, including 23% who started infliximab, 35% who started vedolizumab, and 45% who either initiated or had an adjustment in steroid dosage.

The strengths of our study include the large cohort of patients undergoing evaluation for irColitis, with uniform evaluation and management by a single provider. To our knowledge, this is the first study to evaluate the use of office-based flexible sigmoidoscopy, which offers significant advantages in terms of accessibility as well as safety, for example among patients with multiple comorbidities who may be at high risk or have contraindications for anesthesia. We also show the feasibility of a rapid evaluation program, which requires close communication between oncologists and GI specialists to optimize care delivery. Finally, we provide additional evidence to support the use of sigmoidoscopy as the preferred initial evaluation for patients with suspected irColitis, given its safety, tolerability and high diagnostic yield.

The primary limitation of our study is the focus on a specialized practice at a tertiary care cancer center, which may limit the generalizability to other practice settings. All sigmoidoscopies were performed in an outpatient setting; therefore, patients who were evaluated as inpatients and may have had more severe disease, were excluded from this population. Additionally, as the focus of this study was patients undergoing endoscopic evaluation as part of RAPID-GI, we do not have a comparator group of patients who underwent GI evaluation without endoscopy to isolate the impact of GI specialty evaluation from the results of the endoscopy on patient care. Similarly, we also lack a comparison group of patients who underwent endoscopic evaluation under anesthesia, rather than office-based evaluation. Future directions include evaluating the feasibility of implementing this model in other hospital systems and community-based practices, as well as comparing the time to evaluation between this model and standard endoscopy pathways. Further studies should also assess the adequacy of flexible sigmoidoscopy compared with full colonoscopy in similar patient populations, and evaluate its cost-effectiveness.

In summary, office-based flexible sigmoidoscopy is a safe and effective tool for the rapid diagnosis of irColitis and exclusion of alternative etiologies of symptoms, with an immediate impact on patient care and management plans. Given its high diagnostic yield and rapid availability, without the need for sedation and full bowel preparation, flexible sigmoidoscopy should be considered as a first-line approach for diagnosis in patients with suspected irColitis. Dedicated programs to facilitate the rapid evaluation of patients with suspected irColitis are feasible and can enhance patient outcomes.