Francis, Jasmine H.

Abstract Among 64 patients with Erdheim-Chester disease treated with a BRAF inhibitor (median follow-up 4 years), we found high response rates (85%) but frequent discontinuations (61%), primarily because of adverse events. Additionally, patients experienced poor health-related quality of life and high symptom burden.

Purpose: To analyze the accuracy and thoroughness of three large language models (LLMs) to produce information for providers about immune checkpoint inhibitor ocular toxicities. Methods: Eight questions were created about the general definition of checkpoint inhibitors, their mechanism of action, ocular toxicities, and toxicity management. All were inputted into ChatGPT 4.0, Bard, and LLaMA programs. Using the six-point Likert scale for accuracy and completeness, four ophthalmologists who routinely treat ocular toxicities of immunotherapy agents rated the LLMs answers. Analysis of variance testing was used to assess significant differences among the three LLMs and a post hoc pairwise t-test. Fleiss kappa values were calculated to account for interrater variability. Results: ChatGPT responses were rated with an average of 4.59 for accuracy and 4.09 for completeness; Bard answers were rated 4.59 and 4.19; LLaMA results were rated 4.38 and 4.03. The three LLMs did not significantly differ in accuracy (P = 0.47) nor completeness (P = 0.86). Fleiss kappa values were found to be poor for both accuracy (−0.03) and completeness (0.01). Conclusion: All three LLMs provided highly accurate and complete responses to questions centered on immune checkpoint inhibitor ocular toxicities and management. Further studies are needed to assess specific immune checkpoint inhibitor agents and the accuracy and completeness of updated versions of LLMs.

Purpose: The purpose of this study was to assess the accuracy and completeness of 3 large language models (LLMs) to generate information about antibody–drug conjugate (ADC)-associated ocular toxicities. Methods: There were 22 questions about ADCs, tisotumab vedotin, and mirvetuximab soravtansine that were developed and input into ChatGPT 4.0, Bard, and LLaMa. Answers were rated by 4 ocular toxicity experts using standardized 6-point Likert scales on accuracy and completeness. ANOVA tests were conducted for comparison between the 3 subgroups, followed by pairwise t-tests. Interrater variability was assessed with Fleiss kappa tests. Results: The mean accuracy score was 4.62 (SD 0.89) for ChatGPT, 4.77 (SD 0.90) for Bard, and 4.41 (SD 1.09) for LLaMA. Both ChatGPT (P = 0.03) and Bard (P = 0.003) scored significantly better for accuracy when compared with LLaMA. The mean completeness score was 4.43 (SD 0.91) for ChatGPT, 4.57 (SD 0.93) for Bard, and 4.42 (SD 0.99) for LLaMA. There were no significant differences in completeness scores between groups. Fleiss kappa assessment for interrater variability was good (0.74) for accuracy and fair (0.31) for completeness. Conclusions: All 3 LLMs had relatively high accuracy and completeness ratings, showing LLMs are able to provide sufficient answers for niche topics of ophthalmology. Our results indicate that ChatGPT and Bard may be slightly better at providing more accurate answers than LLaMA. As further research and treatment plans are developed for ADC-associated ocular toxicities, these LLMs should be reassessed to see if they provide complete and accurate answers that remain in line with current medical knowledge.

Summary Histiocytic neoplasms are diverse clonal haematopoietic disorders, and clinical disease is mediated by tumorous infiltration as well as uncontrolled systemic inflammation. Individual subtypes include Langerhans cell histiocytosis (LCH), Rosai–Dorfman–Destombes disease (RDD) and Erdheim–Chester disease (ECD), and these have been characterized with respect to clinical phenotypes, driver mutations and treatment paradigms. Less is known about patients with mixed histiocytic neoplasms (MXH), that is two or more coexisting disorders. This international collaboration examined patients with biopsy‐proven MXH with respect to component disease subtypes, oncogenic driver mutations and responses to conventional (chemotherapeutic or immunosuppressive) versus targeted (BRAF or MEK inhibitor) therapies. Twenty‐seven patients were studied with ECD/LCH (19/27), ECD/RDD (6/27), RDD/LCH (1/27) and ECD/RDD/LCH (1/27). Mutations previously undescribed in MXH were identified, including KRAS, MAP2K2, MAPK3, non‐V600‐BRAF, RAF1 and a BICD2‐BRAF fusion. A repeated‐measure generalized estimating equation demonstrated that targeted treatment was statistically significantly (1) more likely to result in a complete response (CR), partial response (PR) or stable disease (SD) (odds ratio [OR]: 17.34, 95% CI: 2.19–137.00, p = 0.007), and (2) less likely to result in progression (OR: 0.08, 95% CI: 0.03–0.23, p < 0.0001). Histiocytic neoplasms represent an entity with underappreciated clinical and molecular diversity, poor responsiveness to conventional therapy and exquisite sensitivity to targeted therapy.

Purpose: The discovery of MAPK/ERK pathway mutations has led to the approval of MEK-inhibitor (MEKi) cobimetinib in histiocytic neoplasms. However, the pivotal clinical trial that led to the approval of cobimetinib included only 4 patients with Langerhans cell histiocytosis (LCH). Therefore, there is a need to define outcomes with MEKi therapy in patients with LCH. We undertook a multi-center collaboration through the Adult LCH Working Group for the Histiocyte Society to address this gap in knowledge. Methods: Adults with LCH who received MEKi therapy were included from five institutions. Objective response rate (ORR) was evaluated using PET/CT criteria, and CT/MRI if PET scan was not available; response categories were categorized into complete response (CR), partial response (PR), stable disease (SD), and progressive disease (PD). Progression-free (PFS) and overall survival (OS) were calculated from the time of targeted therapy initiation. Risk organ disease was defined as liver, spleen, or bone marrow involvement. Results: 25 adults with LCH were included in this study; median age at MEKi initiation 43y (range, 18-85). Most common systems involved at diagnosis were bone (52%), pituitary (44%), and lung (36%). Risk organ involvement was observed in 7 (28%, 6 liver, 1 liver and spleen) and 2 (8%) patients had intra-axial CNS involvement. Median lines of prior therapy received were 2 (range: 1-5), and 4 (16%) patients received MEKi in first-line. MAPK/ERK pathway status was assessed via target capture next generation sequencing (NGS) in 22 (96%) cases. Amongst those who underwent NGS, alterations identified included: BRAF V600E (2/22, 9%), BRAF deletions (11/22 50%), MAP2K1 or MEK mutations (4/22, 18%), and MAPK1 or ERK2 (n=1, 5%). Overall, 5 patients also received BRAF inhibitors (4 patients as first therapy prior to MEKi and 1 concomitantly due to BRAF V600E positive melanoma). Specific MEKi agents utilized included cobimetinib (n=13) and trametinib (n=12). Among the response evaluable patients (n=23), ORR to MEKi treatment was 78% (9 CR, 9 PR, 4 SD, 1 PD). Among the 7 patients with liver involvement, responses were seen in 2 (29%) cases. ORR did not significantly differ according to mutation status among those with NGS data (78% MAPK mutated vs. 100% unmutated; p=1.0). In this subcohort, all of the 4 cases that did not respond to MEKi had liver involvement; 3 had BRAF deletions (2 SD and 1 PD), and one had MAPK1 mutation (SD). Median follow-up duration after MEKi initiation in the entire cohort was 18 months (95% CI 6 - 30). The median PFS and OS were not reached; 2-year PFS and OS were 78.1% and 79.9%, respectively (Figure 1A). The median PFS was much shorter in those with liver involvement (7 months [95% CI 0 - 14]) compared with those without liver disease (not reached [95% CI 19-not reached] P=0.047; Figure 1B). PFS did not differ significantly between MAPK pathway mutated and unmutated patients (p=0.34) Adverse events (AEs) were reported in 36% (n=9) and most commonly included rash (n=5) and fatigue (n=3). Four patients experienced grade 3-4 AEs including rash, fatigue, and thrombocytopenia. Common causes of MEKi discontinuation included AEs (n=3), progression (n=2), and drug holiday (n=2). At last follow-up, 5 patients died and 18 remained on MEKi. Death was attributable to LCH (n=3), concomitant malignancy (n=1), or unknown causes (n=1). Conclusions: In our multi-institutional study of adults with LCH, we found MEKi therapy to be highly efficacious. However, liver involvement correlated with a lack of response and worse PFS, while mutational status did not. Future studies are needed to evaluate optimum treatment for high-risk subgroups of LCH like liver disease.

Introduction The treatment of Erdheim-Chester disease (ECD) has been revolutionized by the discovery of MAPK/ERK pathway mutations, most common being BRAF-V600E (50%). The US-FDA approved single-agent vemurafenib for BRAF-V600-mutated ECD in 2017 based on a phase II trial that included 22 patients with ECD and 4 with Langerhans cell histiocytosis (LCH), resulting in 100% overall response rate (ORR) by FDG-PET criteria (Diamond et al. JAMA Onc 2018). However, adverse events (AEs) to vemurafenib led to treatment interruption and/or dose modification in all patients, and drug discontinuation in 8 (31%) patients. Long-term outcomes and tolerability of BRAF-inhibitor (BRAFi) monotherapy in real world setting are unknown. In this study, we aimed to address these questions utilizing a large cohort of patients with ECD. Methods Registry-based ( NCT03329274) study of patients with ECD who received BRAFi (vemurafenib or dabrafenib) as any line of treatment. The registry captures detailed longitudinal clinical, treatment, and outcomes data through medical record abstraction. For this study, we focused on the events of discontinuation, dose reductions, and AEs to BRAFi therapy. Time-to-event analyses were conducted using Kaplan-Meier method from the date of initiation of BRAFi. Primary endpoint of the study was event-free survival (EFS), defined as time to progression of disease, discontinuation or alteration of therapy due to any cause (including addition of a MEK inhibitor), or death. Secondary endpoints included progression-free survival (PFS), defined as time to progression of disease or death, and overall survival (OS). Results We included 64 patients with ECD (including 7 with mixed ECD/LCH) in this study. Median age at diagnosis was 56y (range, 26-77) and at initiation of BRAFi was 57y (range, 26-77). Of these, 59 (92%) were BRAF-V600E, 2 (3.1%) were BRAF-wildtype, 1 had BRAF-V471F, 1 had BRAF-PICALM fusion, and 1 had unknown BRAF status. The distribution of organ involvement was bone (91%), neurologic (78%), brain parenchyma (62%), cardiovascular (58%), retroperitoneum (58%), skin (25%), and pulmonary (19%). BRAFi therapy was used as first-line in 38 (59.3%) and as subsequent line in the rest (median 1, range 1-7). The most common initial BRAFi used was vemurafenib (80%) followed by dabrafenib (20%). The median starting dose of vemurafenib was 720 mg bid (range 480-960) and end dose was 480mg bid (range, 240-480). The median start dose of dabrafenib was 100mg bid (range 50-150) and end dose was 50mg bid (range 50-100). Initial dose reduction of BRAFi was required due to AEs in 28 of 52 cases (54%) with final dose data available, encompassing cutaneous (43%), generalized symptoms or fatigue (18%), &gt;1 system affected (14%), lab abnormalities (11%), musculoskeletal (7%), and other (7%) toxicities. Of 60 patients with response data, a response was seen in 51 patients or 85% (42 partial response, 9 complete response, 8 stable disease, and 1 progressive disease). The median follow-up duration was 3.9 years (range 0.2-8.8). Single-agent BRAFi therapy was discontinued in 38 (59%) patients who received it initially (vemurafenib 63% vs. dabrafenib 46%, p=0.28). Reasons for discontinuation included AEs (n=20, 53%), addition of MEK-inhibitor (18%) for reducing cutaneous toxicities (n=5) or augment CNS response (n=2), disease progression (n=4, 11%), death (n=3, 8%), or others (n=4, 11%). 17 patients were rechallenged with a BRAFi after initial discontinuation, and 6 (35%) discontinued it again. The median EFS for the entire cohort from first BRAFi treatment was 2.99y (95% CI 1.6-4.8) and from second BRAFi was 2.82y (95% CI 1.19- not reached). The median PFS and OS were not reached, and 4 patients died at last follow-up (3 from ECD). Conclusions Almost 5 years after the FDA approval of vemurafenib for ECD, our real-world study from long-term follow-up of a large cohort of patients highlights high response rates, but at the cost of adverse events leading to dose reductions and treatment discontinuation in a large subset. Our results also highlight the entity of disease progression on BRAFi therapy. Our results underscore the importance of developing therapies that are efficacious and well-tolerated in this incurable disease.