Search

PURPOSE The risk of immune checkpoint inhibitor therapy–related GI adverse events in patients with cancer and inflammatory bowel disease (IBD) has not been well described. We characterized GI adverse events in patients with underlying IBD who received immune checkpoint inhibitors. PATIENTS AND METHODS We performed a multicenter, retrospective study of patients with documented IBD who received immune checkpoint inhibitor therapy between January 2010 and February 2019. Backward selection and multivariate logistic regression were conducted to assess risk of GI adverse events. RESULTS Of the 102 included patients, 17 received therapy targeting cytotoxic T-lymphocyte antigen-4, and 85 received monotherapy targeting programmed cell death 1 or its ligand. Half of the patients had Crohn’s disease, and half had ulcerative colitis. The median time from last active IBD episode to immunotherapy initiation was 5 years (interquartile range, 3-12 years). Forty-three patients were not receiving treatment of IBD. GI adverse events occurred in 42 patients (41%) after a median of 62 days (interquartile range, 33-123 days), a rate higher than that among similar patients without underlying IBD who were treated at centers participating in the study (11%; P < .001). GI events among patients with IBD included grade 3 or 4 diarrhea in 21 patients (21%). Four patients experienced colonic perforation, 2 of whom required surgery. No GI adverse event–related deaths were recorded. Anti–cytotoxic T-lymphocyte antigen-4 therapy was associated with increased risk of GI adverse events on univariable but not multivariable analysis (odds ratio, 3.19; 95% CI, 1.8 to 9.48; P = .037; and odds ratio, 4.72; 95% CI, 0.95 to 23.53; P = .058, respectively). CONCLUSION Preexisting IBD increases the risk of severe GI adverse events in patients treated with immune checkpoint inhibitors.

PURPOSE Immune checkpoint inhibitor (ICI) therapy often is suspended because of immune-mediated diarrhea and colitis (IMDC). We examined the rate of and risk factors for IMDC recurrence after ICI resumption. METHODS This retrospective multicenter study examined patients who resumed ICI therapy after improvement of IMDC between January 2010 and November 2018. Univariable and multivariable logistic regression analyses assessed the association of clinical covariates and IMDC recurrence. RESULTS Of the 167 patients in our analysis, 32 resumed an anti–cytotoxic T-cell lymphocyte-4 (CTLA-4) agent, and 135 an anti–programmed cell death 1 or ligand 1 (PD-1/L1) agent. The median age was 60 years (interquartile range [IQR], 50-69 years). The median duration from IMDC to restart of ICI treatment was 49 days (IQR, 23-136 days). IMDC recurred in 57 patients (34%) overall (44% of those receiving an anti–CTLA-4 and 32% of those receiving an anti–PD-1/L1); 47 of these patients (82%) required immunosuppressive therapy for recurrent IMDC, and all required permanent discontinuation of ICI therapy. The median duration from ICI resumption to IMDC recurrence was 53 days (IQR, 22-138 days). On multivariable logistic regression, patients who received anti–PD-1/L1 therapy at initial IMDC had a higher risk of IMDC recurrence (odds ratio [OR], 3.45; 95% CI, 1.59 to 7.69; P = .002). Risk of IMDC recurrence was higher for patients who required immunosuppression for initial IMDC (OR, 3.22; 95% CI, 1.08 to 9.62; P = .019) or had a longer duration of IMDC symptoms in the initial episode (OR, 1.01; 95% CI, 1.00 to 1.03; P = .031). Risk of IMDC recurrence was lower after resumption of anti–PD-1/L1 therapy than after resumption of anti–CTLA-4 therapy (OR, 0.30; 95% CI, 0.11 to 0.81; P = .019). CONCLUSION One third of patients who resumed ICI treatment after IMDC experienced recurrent IMDC. Recurrence of IMDC was less frequent after resumption of anti–PD-1/L1 than after resumption of anti–CTLA-4.

22 Background: Inflammatory bowel disease (IBD) increases the risk of developing colorectal cancer (CRC), and colitis-associated CRC (CA-CRC) mortality is on the rise. It has been postulated that CA-CRC may be contributing to the increasing prevalence of early-onset CRC (EOCRC) but supportive studies are currently lacking. Molecular and clinical differences between CA-CRC and sporadic-CRC (S-CRC) have been reported, however outcomes for CA-CRC remains unclear. Signet ring cell carcinoma (SRC) is a rare subtype of CRC which is seen at higher frequencies, along with mucinous histology, in both CA-CRC and EOCRC. In this study, we validate the association of SRC and mucinous (SRC/M) histology with CA-CRC and EOCRC, and utilize it to estimate the amount of EOCRC attributable to undiagnosed or subclinical IBD. Methods: A retrospective study was conducted using three independent mCRC patient datasets from MDACC. The mATTACC discovery cohort consisted of 32 IBD- and 425 S-mCRC patients enrolled in a prospective biomarker trial. Validation of tumor histology was completed with a tumor registry (n=1696), excluding the MSI-High samples, and a real-world evidence (RWE) cohort from MDACC containing 269 CA-mCRC and 29,596 S-mCRC patients, was used as our validation cohort. Results: In the mATTACC cohort SRC/M histology was found in 37.5% of CA-mCRC and 11.7% of S-mCRC, showing a strong association between SRC/M and CA-mCRC (OR = 4.54, 95% CI: 2.19-9.43). The RWE cohort confirmed the correlation of SRC/M with CA-mCRC (28.6%) relative to S-mCRC (11.4%) patients (OR = 3.13, 95%CI: 2.39-4.09). An association was found between SRC/M and EOCRC (OR = 1.35; 95% CI: 1.24-1.47). By comparing the prevalence of SRC/M in EOCRC and late-onset CRC and correcting by the proportion of CA-CRC cases with SRC/M histology, we estimate that between 8.28% to 10.15% of EOCRC may attributable to undiagnosed/subclinical IBD. Using the RWE cohort, median overall survival was determined to be lower for CA-mCRC (31m) relative to S-mCRC (39m; p=0.007), yielding a HR of 1.26 (95% CI: 1.06-1.48). CA-mCRC patients with EOCRC (25m) were also found to have significantly worse outcomes than S-mCRC patients (40m) with EOCRC (p=0.0005; HR = 1.61, 95%CI: 1.23-2.11). Within CA-mCRC, patients with SRC or SRC/M histology (21m) had decreased OS compared to mucinous histology (51m), indicating the poor prognosis of SRC in CA-mCRC (p=0.028; HR=0.53, 95% CI: 0.3-0.94). Conclusions: Tumor biology consistent with CA-CRC, including SRC/M histology, may be present in 8.3% – 10.2% of patients with EOCRC without a clinical diagnosis of IBD, and harbors worse outcomes. Although other confounding biology may be underlying this association, recognition of undiagnosed IBD in CRC patients, especially those with metastatic disease, is important as it may impact prognosis and treatment strategies for this high-risk patient population.

2642 Background: Immune-mediated colitis (IMC) may limit immune checkpoint inhibitors (ICI) treatment. Current guidelines recommend consideration of resuming ICI when IMC symptoms subside to ≤ grade 1. We aimed to investigate the effect of maintenance immunosuppressive therapy (IST) on the outcome of IMC in patients who resume ICI therapy. Methods: We retrospectively studied patients who resumed ICI therapy after adequate treatment of IMC from March 2015 to June 2020 at MD Anderson Cancer Center. Relevant demographic, oncologic, and ICI data were collected and analyzed. Univariate logistic regression analysis was conducted to assess risk factors of IMC recurrence. Results: We included 102 patients with a median age of 61 years. 66% were males and 97% were Caucasians. 48 patients (47%) received IST maintenance in conjunction with ICI resumption and 54 patients did not. Symptoms of IMC recurred in 28 patients, 8 (17%) in the concurrent IST group and 20 (37%) in the other group. Compared to no concurrent IST group, patients on concurrent IST were more likely to have received combined ICI regimen (60% vs 41%, p = 0.003) and more initial ICI doses (9 vs 5 doses, p = 0.030). Concurrent IST group had significantly longer ICI treatment duration on resumption (72 vs 62 days, p = 0.023), more ICI resumed doses (5 vs 4 doses, p = 0.038), and lower IMC recurrence (17% vs 37%, p = 0.027). Patient who received more IST doses, both therapeutic and prophylactic, had lower rate of IMC recurrence (OR 0.72, p = 0.012; table). IST maintenance treatment (OR 0.34, p = 0.024) was associated with lower IMC recurrence rate after ICI resumption. Vedolizumab was the predominant IST used. Overall survival was comparable among the two groups (p = 0.934). Conclusions: Concurrent IST treatment with ICI resumption after IMC was associated with significantly lower IMC recurrence and more extended ICI treatment while reserving similar overall survival to patients without IST maintenance therapy. Future prospective randomized trial of concurrent IST is still merited for further clarification.[Table: see text]

TPS2676 Background: Nivolumab is an anti-PD1 monoclonal antibody approved for treatment of an increasing number of solid tumors and hematological malignancies. However, patients (pts) with history of autoimmune disorders are excluded from the majority of clinical trials testing immune-checkpoint inhibitors (ICI) such as anti-PD1/anti-PD-L1 antibodies. Consequently, the risks of flare ups, worsening of pre-existing autoimmune disorders or risk of de-novo immune related adverse events (irAEs) in pts with dysfunctional immune systems and tumor types who otherwise stand to benefit from ICI therapy are largely unknown, posing a challenge for oncologists. We are conducting a phase Ib study to test the hypothesis that nivolumab can be safely administered to pts with varying severity of Dermatomyositis, Systemic Sclerosis, Rheumatoid Arthritis, Systemic Lupus Erythematosus, Inflammatory Bowel Disease, Multiple Sclerosis and other autoimmune disorders (AIM-Nivo). Methods: AIM-Nivo is an open-label, multi-center ongoing phase Ib study with nivolumab 480mg IV every 28 days in pts with autoimmune diseases and advanced malignancies (NCT03816345). The study has autoimmune disease-specific cohorts overseen by a multidisciplinary group of experts. The primary objective is to assess the overall safety and toxicity profile of nivolumab in pts with autoimmune disorders and advanced malignancies. Secondary objectives are to evaluate the antitumor efficacy; the impact of nivolumab on the autoimmune disease severity indices; and to explore potential biomarkers of response, resistance, or toxicity for each of the autoimmune disease-specific cohorts. Key overall inclusion criteria include age ≥18 years, histologically confirmed advanced or metastatic malignancies in which ICI are approved or have shown clinical activity. Key overall exclusion criteria include prior therapy with anti-PD-1/PD-L1 antibodies. Specific eligibility criteria are defined for each disease-specific cohort. For each autoimmune disorder, severity level of the disease as defined by disease-specific severity indices will be assessed, and up to a total of 12 pts will be included in each disease cohort at each severity level (max 36 pts per cohort). Primary endpoints are dose-limiting toxicities, adverse events (AEs) and serious AEs. Continuous monitoring of toxicity will be conducted. Key secondary endpoints are best objective response per RECIST1.1; progression free and overall survival; and cohort specific tumor tissue, blood, and non-tumor tissue-based biomarkers. The AIM-Nivo trial opened in May 2019 and is enrolling pts through the National Cancer Institute Experimental Therapeutics Clinical Trials Network (ETCTN), Early Drug Development Opportunity Program (EDDOP), and Create Access to Targeted Cancer Therapy for Underserved Populations (CATCH-UP) sites. Clinical trial information: NCT03816345.

2643 Background: Immune checkpoint inhibitor (ICI)-mediated colitis (IMC) is a common and serious adverse event. Although small series have described the clinical presentation of IMC, large multicenter series that integrate clinical, endoscopic, and histologic findings are lacking. Methods: We retrospectively assessed patients who received ICI and had endoscopically confirmed IMC from 2010 to 2019. IMC was graded based on the CTCAE version 5.0 criteria. Multivariate logistic regression analyses were conducted to assess factors associated with recurrence of IMC symptoms and long duration of corticosteroids use (> 70 days). Results: 675 patients were included. 387 patients were males (57%). Median age was 63 years. Melanoma was the most common cancer type (327; 48%). Most (365; 54%) patients received CTLA-4 inhibitor ICI, as monotherapy or in combination with PD-(L)1. Median time from ICI therapy to IMC was 62 days. IMC was grade 2 in 335 (50%) patients, Grade 3 in 181 (27%), and grade 4 in 16 (3%). 155 (23%) patients had mucosal ulceration on endoscopy, 91 of them had severe features (deep, large, or multiple ulcers); 336 (50%) patients had non-ulcerative inflammation. The rest had normal endoscopic findings with histologic inflammation. Most patients were admitted to the hospital for management of IMC (405; 60%) and 16 (3%) needed ICU-level of care. Treatment included corticosteroids in 577 (85%) patients (median duration 52 days), TNF inhibitor in 245 (36%), and vedolizumab in 90 (13%). 202 (32%) patients had recurrent IMC after resolution of symptoms. On multivariate logistic regression, factors associated with IMC recurrence and long (> 70 days) duration of corticosteroid therapy were grade of IMC ( p = 0.049), treatment with infliximab or vedolizumab ( p = 0.044), presence of mucosal ulceration ( p = 0.034 ), or features of active histologic inflammation ( p = 0.076). Of note, patients with mucosal ulceration received infliximab or vedolizumab more frequently ( p < 0.001). For patients with grade 2 IMC, mucosal inflammation on endoscopy and delay in performing endoscopy with time from IMC onset to endoscopy more than a month were associated with IMC recurrence and longer duration of corticosteroid use ( p = 0.029 and p < 0.001, respectively). 16 (3%) patients had colonic perforation, 7 of them underwent surgical resection. No IMC-related death occurred. Conclusions: IMC is a clinically significant adverse event that can lead to premature termination of ICI therapy with high rates of hospital admission. Rarely, it results in colonic perforation requiring surgical intervention and ICU admission. Our data suggest that there is a utility of endoscopic and histologic evaluation in the prediction of worse outcomes from IMC. This finding is particularly important for grade 2 IMC as current guidelines do not recommend endoscopic evaluation for this group.

e16677 Background: Nivo is FDA-approved in sorafenib-experienced, advanced HCC. Post-registration data to portray treatment in a real-world setting is lacking. Methods: We describe safety and efficacy of nivo in patients (pts) from 8 centres (USA n = 181, Asia n = 47, Europe n = 5), documenting overall (OS), progression-free survival (PFS), overall response (ORR) and disease control rates (DCR) (RECIST). Results: We analysed 233 pts, mostly cirrhotic (n = 176, 75%) due to hepatitis C (n = 95, 54.0%) with Barcelona Clinic Stage (BCLC) C HCC (n = 178, 76.4%), Child-Pugh class (CP) A (n = 158, 67.8%) or B (n = 75, 32.2%), and AFP > 400 IU/mL (n = 132, 56.7%). Nivo was given as first (1L, n = 85, 36.5%), second line (2L, n = 130, 55.8%) or > 2L (n = 18, 7.7%), after local (n = 191, 82%) and systemic therapy (n = 148, 63.5%), mostly sorafenib (n = 142, 60.9%). Median duration of nivo was 6.0 months (mo, interquartile range [IQR] 2.6-11.9) and stopped due to progression (n = 109, 46.8%) or toxicity (n = 8, 3.4%). After median follow up of 7 mo (IQR 3.0-12.3), ORR was 22.4% and DCR was 52.1%. Best responses (n = 219, 94%) included complete and partial responses in 18 (7.7%) and 31 (13.3%) pts respectively, stable disease in 65 (27.9%) and progressive disease in 105 (45.1%), not dissimilar by CP (p = 0.26). Median OS was 12.2 mo (95%CI 8.4-16.0) and predicted by CP (CPA 16.3 mo 95%CI 11.7-20.8; CPB 7.3 mo 95%CI 4.2-10.4; hazard ratio [HR] 1.9, p = 0.01), nivo line (1L 16.3 mo 95%CI 8.0-24.5; > 1L 10.4 mo 95%CI 7.4-13.5; HR 0.68, p = 0.05), and PVT (PVT- 13.8 mo 95%CI 11.7-16.0; PVT+ 10.4 mo 95%CI 7.8-13.0; HR 1.8, p = 0.015) but not cirrhosis, AFP, BCLC or steroid use (p > 0.05). Median PFS was 10.1 mo (95%CI 6.1-14.2), predicted by BCLC (A-B 19.0 mo 95%CI 7.1-30.8; C 8.2 mo 95%CI 4.9-11.4; HR 2.8, p = 0.002) and line (1L 18.2 mo 95%CI 10.4-25.9; > 1L 8.2 mo 95% CI 6.2-10.2; HR 0.60, p = 0.021), but not cirrhosis, AFP, or steroid use (p > 0.05). 26 pts (11.2%) suffered > = Grade 2 toxicities, most commonly fatigue (n = 29, 24.7%). Conclusions: Real-world use of nivo in advanced HCC across line of therapy suggests reproducible clinical efficacy and safety compared to prospective trials.

e16500 Background: Eosinophilic colitis (EoC) is a rare form of eosinophilic gastrointestinal disease characterized by diffuse eosinophilic infiltration in the deep lamina propria and extensive crypt epithelial injury with apoptosis and crypt abscesses. The pathophysiology is still unclear and associated with many etiologies. We described the clinical characteristics of EoC in cancer patients. Methods: We retrospectively reviewed the pathology report of the colonic samples from endoscopy or surgical resection obtained between 01/2000 and 12/2018 at MD Anderson Cancer Center. The included cases had significant eosinophilia with key word description of “numerous”, “many” or “dominant” on the pathology. Clinical data were collected and analyzed. Results: A total of 41 cases were included based on the inclusion criteria with 34% men, with a median age of 60. 27% had NSAID use and 40% were smokers. One-fourth had co-existing autoimmune conditions. 78% had cancer diagnosis (hematologic at 41%, followed by genitourinary and gastrointestinal origin) with the rest carrying no cancer diagnosis. Half of the cases received chemotherapy with median duration of 180 days between chemotherapy to EoC onset. Common EoC symptoms included diarrhea, and abdominal pain which were present in 76% patients. Diarrhea was more prevalent in patients who received chemotherapy (85% vs 42%). Median duration of EoC symptoms was shorter in cancer patients (240 vs 30 days) and chemotherapy exposure (120 vs 30 days). Hospitalization for EoC was required in 15% cases. Most commonly administered treatment for EoC was immunosuppressants (22%). Normal colonoscopy was reported in 88% patients at the time of EoC diagnosis. All-cause mortality was 37% and was mostly related to underlying malignancy. Conclusions: The disease course of EoC in cancer patients on chemotherapy is not well defined. It appeared to be less severe than non-cancer patients, which raised the possibility of two separate subclasses of this disease entity. Cancer/chemotherapy related EoC could be a GI specific paraneoplastic manifestation, and behave differently from other etiology. [Table: see text]

3067 Background: ICIs are efficacious treatment for several advanced malignancies. IMC can limit their use, and can be refractory to medical treatment (immunosuppression) with significant morbidity. Gut microbiome alteration affects IMC development. We sought FMT as a novel therapy for IMC refractory to immunosuppressive therapy. Methods: 15 patients who received FMT for IMC after failure of immunosuppressive therapy were included (6/2017-1/2020). FMT was performed via colonoscopy with healthy donor’s stool. Results: Median age was 55 years with 67% males. 5 patients received PD(L)-1, one CTLA-4 and 9 on combination. Majority had genitourinary cancers followed by melanoma. Median time from ICI to IMC was 75 days. 14 patients had grade 3-4 diarrhea and 9 had grade 3-4 colitis. Endoscopy showed mucosal inflammation in 12 patients and normal mucosa in 3 patients. IMC was refractory to 2-3 doses of infliximab or vedolizumab after corticosteroids prior to FMT. Median time from IMC onset to FMT was 75 days. 13 patients received one . 11 patients achieved clinical response within 10 days of FMT (7-14). Symptom remission was maintained for a median follow-up of 13 months. 6 patients resumed non ICI cancer treatments after FMT. 4 patients had persistent symptoms; 2 continued on vedolizumab, 1 had total colectomy, and 1 transferred to hospice. 4-8 weeks after FMT, endoscopic remission was achieved in 64% of the 11 patients who responded to FMT. No adverse events were reported. Conclusions: FMT treatment was successful in 73% of patients with for IMC refractory to immunosuppressive therapy. Controlled clinical trials are warranted to confirm our conclusion. [Table: see text]

TPS3158 Background: Immune checkpoint inhibitors (ICI) such as anti-PD-1/PD-L1 antibodies have become a pivotal approach to cancer therapy. Nivolumab is an anti-PD1 antibody approved for an increasing number of solid tumors and hematological malignancies. However, patients (pts) with history of autoimmune disorders are excluded from the majority of clinical trials testing ICI. Consequently, the risks of flares and worsening of pre-existing autoimmune disorders in pts with tumor types who otherwise stand to benefit from ICI therapy are largely unknown, posing a challenge for oncologists. We are conducting a phase Ib study to test the hypothesis that nivolumab can be safely administered to pts with varying severity of Dermatomyositis, Systemic Sclerosis, Rheumatoid Arthritis, Systemic Lupus Erythematosus, Inflammatory Bowel Disease, Multiple Sclerosis, Sjögren's Syndrome and Other Autoimmune Disorders (AIM-Nivo). Methods: AIM-Nivo is an open-label, multi-center ongoing phase Ib study with nivolumab 480mg IV every 28 days in pts with autoimmune diseases and advanced malignancies (NCT03816345). The study has autoimmune disease-specific cohorts overseen by a multidisciplinary group of experts. The primary objective is to assess the overall safety and toxicity profile of nivolumab in pts with autoimmune disorders and advanced malignancies. Secondary objectives are to evaluate the antitumor efficacy, the impact of nivolumab on the autoimmune disease severity indices, and to explore potential biomarkers of response, resistance or toxicity. Key overall inclusion criteria include age ≥18 years, histologically confirmed advanced malignancies in which ICI are approved or have shown clinical activity. Key overall exclusion criteria include prior therapy with anti-PD-1/PD-L1 antibodies. Specific eligibility criteria are defined for each disease-specific cohort. For each autoimmune disorder, severity level of the disease as defined by disease-specific severity indices will be assessed, and up to a total of 12 pts will be included in each disease cohort at each severity level. Primary endpoints are dose-limiting toxicities, adverse events (AEs) and serious AEs. Continuous monitoring of toxicity will be conducted. Key secondary endpoints are best objective response, progression free and overall survival and cohort specific tumor tissue, blood and non-tumor tissue-based biomarkers. The AIM-Nivo trial opened in May 2019 and is enrolling pts through the National Cancer Institute Experimental Therapeutics Clinical Trials Network (ETCTN). Clinical trial information: NCT03816345 .

e24048 Background: Cancer patients are at a significantly increased risk of rCDI by virtue of a compromised immune function from underlying malignancy, anti-cancer therapy and frequent antibiotic use for opportunistic infections. Furthermore, cancer patients have shown to have a lower response rate to standard oral antibiotics for CDI. Data is limited in regards to the safety and efficacy of FMT in managing rCDI in cancer patients. We aim to describe our experience of the same at a tertiary care cancer center. Methods: We conducted a retrospective study of cancer patients who underwent FMT for rCDI at the MD Anderson Cancer Center 06/2017-01/2020. FMT was performed through colonoscopy with universal donors’ stool. Baseline clinical data were collected and analyzed. Results: Our sample comprises 19 patients of whom 12 had solid tumors and 7 hematological malignancies, most of which were stage 4 at the time of FMT. The mean age was 66.5 years. Most patients received proton pump inhibitor, antibiotics and cancer therapy within 3-6 months of FMT. On average, each patient had 3 episodes of CDI, received 4 courses of antibiotic treatments, and required 4 CDI related hospitalizations within 1 year prior to FMT. Majority of the CDI episodes were managed with a combination of antibiotics. Bezlotoxumab was used in 4 cases. 18 patients received FMT once, while one patient was treated with FMT thrice. Clinical response was seen in 74% patients with a median time of 1 day. 5 patients had refractory CDI including 3 recurrent rCDI, and 2 persistent symptoms. Compared to patients with good response, these refractory cases received more frequent antibiotics following FMT (100% vs 43%, p=0.033). Side effects were mostly mild GI complaints in 15.8% patients with no serious adverse events or mortality related to FMT. Conclusions: Our study shows that FMT is a safe and effective treatment for rCDI in cancer patients and provides rapid resolution of symptoms; Subsequent antibiotic use for the management of cancer related complications can negatively affect the efficacy of FMT. [Table: see text]

TPS2088 Background: ICIs have become the therapeutic standard for many cancers but are associated with unique and diverse irAEs that often occur at home. Appropriately timed and specific interventions are critical to recovery. Thus, there is a need to effectively & efficiently monitor in real time pts treated with ICIs. To improve outcomes, we have activated a clinical trial developed to determine the feasibility and safety of an electronically enabled strategy to remotely monitor symptoms and prompt communication that will guide and inform specific patient-driven “course corrections” in response to potential irAEs. Methods: This is an adaptive prospective trial that uses a mobile irAE-specific PRO application we developed to monitor and alert the care team in real time when severe symptoms are reported. In parallel with the mobile symptom collection, serum and urine biomarkers are collected at baseline, first tumor restaging, and upon the development of irAEs. Optional stool microbiome analyses are also performed. To facilitate the generalizability of our inferences, we are using broad inclusion criteria: ECOG performance status ≤3; any line of ICI given as standard of care or as part of therapeutic clinical trials; elderly pts are included. Because the relationship between PROs and irAEs is currently undefined, we designed our trial to use adaptive symptom thresholds that will notify the healthcare team of suspicion for irAEs. The mobile application will use these dynamic thresholds to determine whether or not to alert the healthcare team. The positive and negative predictive value of each symptom for identifying subsequent irAEs will be assessed at scheduled interim analysis time points. The care teams’ responses to the alerts, and all of the clinical outcomes for the pts over time will be collected as part of the trial. The primary goal of the trial is the assessment of the predictive power of the mobile PRO symptom collection in combination with serum and urine markers to identify grade 2 or higher adverse events that require intervention (e.g., dose modifications, hospitalizations, and therapeutic interventions) within two weeks of symptom onset. Effective remote monitoring of irAEs will leverage our understanding of ICI toxicity and empower pts to be effective partners in their care. The trial has currently enrolled 17 pts towards the enrollment target of 100 pts. Clinical trial information: PA19-0095 .

e15102 Background: Immune checkpoint inhibitors (ICIs) are efficacious in treating many advanced malignancies. However, drug induced colitis limits their use significantly. We present cases with chronic refractory ICI colitis requiring long term immunosuppressive therapy and favorable cancer outcomes. Methods: We identified 3 cases who developed ICI colitis and persisted for longer than 6 months requiring long term immunosuppressant therapy. The patients’ clinical data was collected. Results: All patients were male, with median age of 55. Two had melanoma, one urothelial cancer. All patients received PD-1(L)-1 agents with colitis onset about 3-6 months after ICI treatment. All patients were taken off ICI permanently due to colitis over a duration of 19-30 months (peak grade of 3). Endoscopy of all three were grossly unremarkable, with lymphocytic colitis on pathology in all three. Patient A’s colitis achieved clinical and histological remission after 5 doses of vedolizumab, however recurred after 9 months with biopsy proven recurrence. Current management is anti-diarrheal medication given mild symptom. Patient B had more aggressive disease, with first recurrence after 2 doses of infliximab and steroid, achieved initial histological remission after 3 doses of vedolizumab and 2 additional infliximab, then had 2nd recurrence after 3 months. Fecal microbiota transplantation as compassionate treatment was not effective. Thereafter, patient was resumed on vedolizumab ever since with clinical remission. Last histology evaluation showed persistent lymphocytic colitis 30 months after initial diagnosis. The third case had initial histological remission after 3 doses of vedolizumab, followed by recurrence after 4 months, which triggered another 3 doses of vedolizumab, with persistent grade 3 diarrhea despite resolution of histological inflammation after total 6 doses of vedolizumab. Patient was subsequently treated with steroid and ustekinumab with clinical remission. At the end of follow up, all three patients have sustained cancer remission over 19-31 months. Conclusions: IMC is a common adverse event from ICI, and could progress to a chronic inflammatory condition that require long term treatment. Histological remission does not preclude future recurrence of colitis. Persistent toxicity could be a surrogate marker for enduring ICI effect which may be associated with favorable cancer outcome. Long term immunosuppression can be essential for refractory ICI colitis cases for clinical remission.

e15166 Background: ICI therapies have revolutionized the landscape of cancer treatment. However, their increased use has also contributed to adverse events, e.g. IMDC. Clinically it is difficult to distinguish between infectious etiology or IMDC. It is unclear whether antimicrobial treatment substantially impacts the disease course. Here, we evaluated the characteristics and outcome of IMDC in cancer patients with superimposed diarrheagenic GI infections. Methods: We retrospectively evaluated cancer patients who received ICI with clinical symptoms of IMDC and confirmed stool microbiology of E. coli or non-CMV viral infections either at the time of IMDC, or within 60 days after IMDC diagnosis at MD Anderson Cancer Center 01/2011-08/2019. We described the disease course and outcome of IMDC based on their status of GI infection and antimicrobial treatment. Results: Total 72 patients were included, among them, 50 control patients with IMDC and no GI infection, 22 had diarrheagenic infections, composed of 17 E coli pathogens: Enteropathogenic, Enterotoxigenic, Enteroaggregative, and E coli 0157:H7 serotype; and 5 viral etiologies: Adenovirus, Norovirus, and Sapovirus. Patients in the infection group had higher grade of colitis (42.9% vs 18.4%, P = 0.041), and more frequent hospitalization (86.4% vs 62%, P = 0.052). 68.2% patients in the infection group received infliximab/vedolizumab add-on treatment compared to 40% in the no infection group (P = 0.078). Patients with GI infection and antimicrobial treatment had much higher IMDC recurrence rate than no antimicrobial treatment (50% vs 0, P = 0.015). GI infection was not associated with higher IMDC recurrence or worse overall survival of these patients. Conclusions: IMDC can occasionally be complicated by infection from common GI pathogens. Antimicrobial treatment was mostly used in severe IMDC cases and did not circumvent the need for immunosuppressant or improve the clinical outcomes. GI infection was not associated with higher IMDC recurrence or worse overall survival of these patients. [Table: see text]

531 Background: The impact of corticosteroid treatment (CT) on the efficacy of immune checkpoint inhibitors (ICI) in hepatocellular carcinoma (HCC) is undefined. We evaluated whether CT administered at baseline (bCT) or concurrently to ICI (cCT) influences clinical outcomes of HCC patients treated with ICI. Methods: This retrospective, multi-center observational study was conducted across 9 tertiary academic referral centers collected 341 HCC patients who received ICI across 3 continents between January 1, 2016 and April 1, 2019. Outcome measures included overall (OS) and progression-free survival (PFS) calculated from time of ICI commencement and overall response rates (ORR) defined by Response Evaluation Criteria in Solid Tumors (v1.1) on 6-8 weekly periodic restaging. Results: Of 331 eligible patients, 254 (76%) had BCLC-C stage HCC and received mostly PD(L)-1 ICI monotherapy (n=250, 85%). Median OS was 12.1 months (95%CI 9.2-15.0 months) and median PFS was 8.1 months (95%CI 6.3-10 months). In total 81 patients (24%) received >10 mg prednisone equivalent daily either as bCT (n=15, 4%) or cCT (n=66, 20%). Indications for CT included procedure/prophylaxis (n=37, 45%), management of irAE (n=31, 37%), cancer-related symptoms (n=5, 2%) or comorbidities (n=8, 3%). Neither overall CT, bCT nor cCT predicted for worse OS, PFS nor ORR in uni- and multi-variable analyses (p>0.05). CT for cancer-related indications predicted for shorter PFS (2.4 vs. 11.3 months, p=0.01), OS (4.5 vs. 12.8 months, p=0.05) and reduced ORR (p=0.03) compared to cancer-unrelated indications. Conclusions: This is the first study to demonstrate that neither bCT nor cCT appear to influence response and OS following ICI in HCC. Worse survival and ORR in CT recipients for cancer-related indications appears driven by the poor prognosis associated with symptomatic HCC.

1541 Background: In our clinical practice at a tertiary cancer center, we have observed increased adenoma detection rate (ADR) in patients with breast cancer. Here, we describe ADR in patients with breast cancer to define the appropriate timing to initiate colonoscopy screening in these patients. Methods: We conducted a retrospective study of patients with breast cancer who underwent a colonoscopy after their diagnosis of breast cancer between 2000 and 2017. A control group ( n = 3295) comprised patients without any type of cancer who underwent their first screening colonoscopy between 2008 and 2017 was used in the logistic regression. Results: Of the 62,820 patients who had a diagnosis of breast cancer, 3304 were included. The mean age was 59 years. Regarding ADR, 1803 patients (55%) had adenomas. High-grade dysplasia was evident in 28% of polyps and invasive adenocarcinoma in 172 (5%). The median time from breast cancer diagnosis to adenoma detection was 3 years (IQR 1-6). The ADR was 21% in patients younger than 40 years ( n=63), 39% in patients between 40 and 50 years ( n=314), 54% in patients between 50 and 60 years ( n=1420), and 60% in patients older than 60 years ( n=1507). ADR in patients younger than 50 years of age who do not have a family history of colorectal cancer or a body mass index (BMI) higher than 30 kg/m2 was 26%. A subsequent colonoscopy was performed in 831 patients who had colonic adenoma in the initial colonoscopy. The ADR was 40% in patients who had a repeat colonoscopy within 3 years, 50% within 3-5 years, and 53% > 5 years. Multivariate logistic regression analyses revealed an increased risk of colon adenoma with older age, male sex, higher BMI, and personal history of breast cancer ( P

2616 Background: A deep understanding of the immunobiology of checkpoint inhibitor (CPI) induced immune related toxicities, such as immune related enterocolitis (irEC), and how these compare to the immune signatures in tumors could lead to the development of strategies that de-couple autoimmunity from anti-tumor immunity. Methods: Total RNA from patient-matched irEC and normal colon FFPE tissue from patients [n = 12] receiving CPIs were profiled with the 770 gene NanoString nCounter PanCancer Immune Profiling Panel (NanoPCIP). The mean fold change in gene expression from normal vs. irEC inflamed colonic tissue and baseline vs. on-treatment tumor samples from patients responding or non-responding to ipilimumab based therapy were analyzed. C57BL/6 mice with B16.BL6 melanoma tumors were treated with systemic anti-IL-6 + anti-CTLA-4 vs. anti-CTLA4 alone vs. placebo and tumor size was measured. Results: In patients with irEC, the highest significantly upregulated differentially expressed gene (DEG) in inflamed colon tissue encoded for IL-6 (Fold change +24.1). None of the significant and highest upregulated DEGs in the colitis, including IL-6, were significantly upregulated in responding tumors. Interestingly, IL-6 was also the highest upregulated DEG in non-responding tumors numerically. When comparing mean fold changes across these analyses, the gene with the largest difference in upregulatation between colitis and responding tumors was IL-6; the other highest upregulated genes in colitis encoded for neutrophil and monocyte chemotactic molecules. In our mouse models, the addition of IL-6 blockade to anti-CTLA-4 therapy significantly improved tumor shrinkage compared to anti-CTLA-4 alone. Conclusions: Our data demonstrates that IL-6-mediated inflammation may be more prevalent in irEC and tumors not responding to CPIs than in tumors responding, and blocking IL-6 enhances CPI anti-melanoma activity. Targeting IL-6 may ameliorate irEC without hindering anti-tumor immunity.

2571 Background: Current treatment guidelines for immune-mediated colitis (IMC) recommend 4 to 6 weeks of steroids as first-line therapy, followed by nonsteroidal immunosuppressive therapy (NSIST) (infliximab or vedolizumab) in patients who do not respond to steroids. We assessed the effect of early NSIST introduction and number of NSIST infusions on clinical outcomes. Methods: We performed a retrospective review of patients with IMC who received NSIST between January and December 2018. Logistic regression analyses were used to assess associations between clinical features and outcomes of IMC (Table). Results: Of the 1,459 patients who received immune checkpoint inhibitor, 179 developed IMC of any grade; 84 of them received NSIST. Of the 84 patients who received NSIST, 79% were males with mean age of 60. Compared with patients who received NSIST >10 days after IMC onset, patients who received early NSIST (≤10 days) required fewer hospitalizations ( P=0.03), experienced steroid taper failure less frequently ( P=0.03), had fewer steroid tapering attempts ( P

2577 Background: Immune checkpoint inhibitor (ICI) therapy is often suspended because of immune-mediated diarrhea and colitis (IMDC). We examined the recurrence rate and risk factors for IMDC after ICI resumption. Methods: This retrospective multicenter study examined patients who resumed ICI therapy after improvement of IMDC between 1/2010 and 11/2018. Univariate and multivariate logistic regression analyses assessed the association of clinical covariates and IMDC recurrence. Results: Of the 167 patients in our analysis, 32 resumed an anti-CTLA-4 agent and 135 an anti-PD-1/L1 agent. The median duration from IMDC to restart of ICI treatment was 49 days (IQR, 23-136). IMDC recurred in 57 (34%) patients overall (44% of those resuming an anti–CTLA-4 and 32% resuming an anti–PD-1/L1 agent); 47 of these patients (82%) required immunosuppressive therapy for recurrent IMDC (Table). The median duration from ICI resumption to IMDC recurrence was 53 days (IQR 22-138). On multivariate logistic regression, patients who received anti-PD-1/L1 therapy at initial IMDC had a higher risk of IMDC recurrence (odds ratio [OR], 3.45, 95%CI, 1.59-7.69; P

e13070 Background: In the last two decades, the incidence of colorectal cancer (CRC) has decreased dramatically after the implementation of CRC screening in the United States. Several risk factors for colonic adenoma (CA), the main precursor for CRC, have been found. Whether personal history of non-colorectal cancer (NCRC) is a risk factor for CA has not been studied. Here, we assess the risk of CA in patients with NCRCs. Methods: We conducted a retrospective study of cancer patients who underwent colonoscopy after cancer diagnosis between 2009 and 2018. We included patients without history of NCRC as a control group. Multivariate logistic regression was used to assess independent risk factors for CA (Table 1). Results: Total of 9408 patients with NCRC were included; CA was detected in 4503 (48%). Histology revealed tubulovillous features in 611 (14%) patients and villous in 51 (1%). High grade dysplasia was detected in 1,317 (29%) patients and adenocarcinoma in 388 (9%). The rate of adenocarcinoma was the highest in patients with multiple myeloma (14%). Adenoma detection rate (ADR) was 30% in patients younger than 40 years ( n= 1621), 32% in patients between 40 and 50 years ( n= 812), 47% in patients between 50 and 60 years ( n= 2892), and 55% in patients older than 60 years ( n= 4493). Multivariate analysis revealed an increased risk of CA with old age, male sex, family history of CRC, and high body mass index ( P< 0.05). The median time from NCRC diagnosis to CA detection was 3 years (IQR 1-8). Conclusions: ADR in patients with a personal history of NCRC is higher than the ADR of patients without NCRC. CRC screening should be performed after the diagnosis of NCRC is made, even if it was before the standard threshold of CRC screening age of 50 years.[Table: see text]

10063Background: Immune checkpoint inhibitors (ICPIs) are gaining more popularity as treatment for advanced cancers. However, immune-related adverse events (irAE) might limit their use. Immunosuppr...

689 Background: Recent reports suggest colorectal cancer incidence is steadily increasing in adults < 50 years old. This population has a number of unique pre-disposing conditions, such as inflammatory bowel disease (IBD), that may alter the disease biology within these patients. Methods: Using a cohort of 1876 patients with metastatic CRC (mCRC), we identified 622 patients aged < 50 years and reviewed their medical records for a diagnosis of IBD. Clinical, pathologic and overall survival (OS) were compared between patients with and without IBD in this age group. Results: Twenty of 622 (3.2%) early-onset mCRC patients had IBD. These patients were more likely to have mucinous or signet ring histologic features (50.0% vs 15.6%, OR 5.43, 95% CI 2.20-13.41, P < 0.0001) and less likely to have APC mutations (15.0% vs 42.7%, P = 0.019). A trend toward an increased prevalence of right-sided location (40.0% vs 24.1%, P = 0.053) and TP53 mutations (85.0% vs 65.1%, P = 0.091) and decreased prevalence of PIK3CA mutations (0.0% vs 15.5%, P = 0.057) and KRAS mutations (35.0% vs 50.0%, P = 0.11) were noted. There were no differences in prevalence of BRAF V600 (P = 0.99), CTNNB1 (P = 0.99), NRAS (P = 0.99), PTEN (P = 0.43), or SMAD4 (P = 0.75) mutations between patients with IBD and other early-onset mCRC. The presence of IBD in early-onset mCRC patients was associated with a worse OS compared to patients without IBD but who had mCRC (HR 1.82, 95% CI 0.90-3.69, P = 0.026). Conclusions: Early-onset mCRC patients with IBD are significantly more likely to have mucinous or signet ring histologic features and have fewer mutations in canonical CRC genes such as KRAS, PIK3CA, and APC. This suggests that this population likely develops through a unique molecular pathogenesis and they may respond differently to therapy. Further molecular characterization of this population is required.

9026 Background: TKI258 is an inhibitor of type III to V RTKs including VEGF-Rs, FGF-Rs (FGF-R1, 2 and 3), PDGFR-β, c- KIT and FLT3 with preclinical pharmacology profile consistent with these targe...