Intricacies of Human-AI Interaction in Dynamic Decision-Making for Precision Oncology: A Case Study in Response-Adaptive Radiotherapy.

Background: Adaptive treatment strategies that can dynamically react to individual cancer progression can provide effective personalized care. Longitudinal multi-omics information, paired with an artificially intelligent clinical decision support system (AI-CDSS) can assist clinicians in determining optimal therapeutic options and treatment adaptations. However, AI-CDSS is not perfectly accurate, as such, clinicians' over/under reliance on AI may lead to unintended consequences, ultimately failing to develop optimal strategies. To investigate such collaborative decision-making process, we conducted a Human-AI interaction case study on response-adaptive radiotherapy (RT). Methods: We designed and conducted a two-phase study for two disease sites and two treatment modalities-adaptive RT for non-small cell lung cancer (NSCLC) and adaptive stereotactic body RT for hepatocellular carcinoma (HCC)-in which clinicians were asked to consider mid-treatment modification of the dose per fraction for a number of retrospective cancer patients without AI-support (Unassisted Phase) and with AI-assistance (AI-assisted Phase). The AI-CDSS graphically presented trade-offs in tumor control and the likelihood of toxicity to organs at risk, provided an optimal recommendation, and associated model uncertainties. In addition, we asked for clinicians' decision confidence level and trust level in individual AI recommendations and encouraged them to provide written remarks. We enrolled 13 evaluators (radiation oncology physicians and residents) from two medical institutions located in two different states, out of which, 4 evaluators volunteered in both NSCLC and HCC studies, resulting in a total of 17 completed evaluations (9 NSCLC, and 8 HCC). To limit the evaluation time to under an hour, we selected 8 treated patients for NSCLC and 9 for HCC, resulting in a total of 144 sets of evaluations (72 from NSCLC and 72 from HCC). Evaluation for each patient consisted of 8 required inputs and 2 optional remarks, resulting in up to a total of 1440 data points. Results: AI-assistance did not homogeneously influence all experts and clinical decisions. From NSCLC cohort, 41 (57%) decisions and from HCC cohort, 34 (47%) decisions were adjusted after AI assistance. Two evaluations (12%) from the NSCLC cohort had zero decision adjustments, while the remaining 15 (88%) evaluations resulted in at least two decision adjustments. Decision adjustment level positively correlated with dissimilarity in decision-making with AI [NSCLC: ρ = 0.53 ( p < 0.001); HCC: ρ = 0.60 ( p < 0.001)] indicating that evaluators adjusted their decision closer towards AI recommendation. Agreement with AI-recommendation positively correlated with AI Trust Level [NSCLC: ρ = 0.59 ( p < 0.001); HCC: ρ = 0.7 ( p < 0.001)] indicating that evaluators followed AI's recommendation if they agreed with that recommendation. The correlation between decision confidence changes and decision adjustment level showed an opposite trend [NSCLC: ρ = -0.24 ( p = 0.045), HCC: ρ = 0.28 ( p = 0.017)] reflecting the difference in behavior due to underlying differences in disease type and treatment modality. Decision confidence positively correlated with the closeness of decisions to the standard of care (NSCLC: 2 Gy/fx; HCC: 10 Gy/fx) indicating that evaluators were generally more confident in prescribing dose fractionations more similar to those used in standard clinical practice. Inter-evaluator agreement increased with AI-assistance indicating that AI-assistance can decrease inter-physician variability. The majority of decisions were adjusted to achieve higher tumor control in NSCLC and lower normal tissue complications in HCC. Analysis of evaluators' remarks indicated concerns for organs at risk and RT outcome estimates as important decision-making factors. Conclusions: Human-AI interaction depends on the complex interrelationship between expert's prior knowledge and preferences, patient's state, disease site, treatment modality, model transparency, and AI's learned behavior and biases. The collaborative decision-making process can be summarized as follows: (i) some clinicians may not believe in an AI system, completely disregarding its recommendation, (ii) some clinicians may believe in the AI system but will critically analyze its recommendations on a case-by-case basis; (iii) when a clinician finds that the AI recommendation indicates the possibility for better outcomes they will adjust their decisions accordingly; and (iv) When a clinician finds that the AI recommendation indicate a worse possible outcome they will disregard it and seek their own alternative approach.

The Current Role of Human Papillomavirus Circulating Tumor DNA in Oropharynx Cancer.

ABSTRACT: Human papillomavirus infection is currently implicated in the majority of oropharyngeal squamous cell carcinoma cases diagnosed in the United States. Circulating tumor DNA (ctDNA) has emerged as a potential biomarker for human papillomavirus-related oropharyngeal squamous cell carcinoma and has the opportunity to improve the diagnosis, treatment, and surveillance of patients with this disease. Changes in ctDNA levels during and after primary therapy may be related to disease response, which can possibly have implications for treatment intensification or de-escalation strategies. Further, ctDNA seems to be sensitive and specific for disease recurrence and may improve upon current methods for assessing both treatment response and failure. In this review, we examine the relevant literature on the use of ctDNA for oropharyngeal cancer treatment and surveillance and discuss current limitations and future directions for this promising biomarker.

Cannabis Use in Patients Seen in an Academic Radiation Oncology Department.

PURPOSE: Cannabis use rates are increasing in the United States. Patients with cancer use cannabis for many reasons, even without high-quality supporting data. This study sought to characterize cannabis use among patients seen in radiation oncology in a state that has legalized adult nonmedical use cannabis and to identify key cannabis-related educational topics. METHODS AND MATERIALS: Cannabis history was documented by providers using a structured template at patient visits in an academic radiation oncology practice October 2020 to November 2021. Cannabis use data, including recency/frequency of use, reason, and mode of administration, were summarized, and logistic regression was used to explore associations between patient and disease characteristics and recent cannabis use. A multivariable model employed stepwise variable selection using the Akaike Information Criterion. RESULTS: Of 3143 patients total, 91 (2.9%) declined to answer cannabis use questions, and 343 (10.9%) endorsed recent use (≤1 month ago), 235 (7.5%) noted nonrecent use (>1 month ago), and 2474 (78.7%) denied history of cannabis use. In multivariable analyses, those ≥50 years old (odds ratio [OR], 0.409; 95% confidence interval [CI], 0.294-0.568; P < .001) or with history of prior courses of radiation (OR, 0.748; 95% CI, 0.572-0.979; P = .034) were less likely, and those with a mental health diagnosis not related to substance use (OR, 1.533; 95% CI, 1.171-2.005; P = .002) or who smoked tobacco (OR, 3.003; 95% CI, 2.098-4.299; P < .001) were more likely to endorse recent cannabis use. Patients reported pain, insomnia, and anxiety as the most common reasons for use. Smoking was the most common mode of administration. CONCLUSIONS: Patients are willing to discuss cannabis use with providers and reported recent cannabis use for a variety of reasons. Younger patients new to oncologic care and those with a history of mental illness or tobacco smoking may benefit most from discussions about cannabis given higher rates of cannabis use in these groups.

The role of adjuvant chemotherapy in the management of acute promyelocytic leukemia differentiation syndrome.

Acute Promyelocytic Leukemia (APL) is characterized by the t(15;17) chromosomal translocation resulting in a PML-RARA fusion protein. The all-trans-retinoic acid (ATRA) and Arsenic Trioxide (ATO) only regimens have demonstrated success in treating low- and intermediate-risk patients. However, induction with ATRA/ATO only regimens have been showing increased incidence of differentiation syndrome (DS), a potentially lethal complication, traditionally treated with dexamethasone. We conducted a three-institution retrospective study, aiming to evaluate the role of short-term adjuvant chemotherapy in managing moderate DS for patients with low- or intermediate-risk APL initially treated with ATRA/ATO only protocols. We evaluated the difference in incidence and duration of moderate DS in APL patients who were treated with ATRA/ATO with or without adjuvant chemotherapy. 57 low- or intermediate-risk APL patients were retrospectively identified and included for this study; 36 patients received ATRA/ATO only induction treatment, and 21 patients received ATRA/ATO/adjuvant chemotherapy combination induction therapy. Similar proportions of patients experienced DS in both groups (66.7% vs. 81.0%, P = 0.246). The median duration of DS resolution in patients receiving ATRA/ATO only was 17 days (n = 23), and in patients receiving combination therapy was 8 days (n = 16) (P = 0.0001). The lengths of hospital stay in patients receiving ATRO/ATO only was 38 days (n = 7), and in patients receiving combination therapy was 14 days (n = 17) (P = 0.0007). In conclusion, adding adjuvant chemotherapy to ATRA/ATO only protocol may reduce the duration of DS and the length of hospital stay during APL induction treatment.

Occurrence of acute myeloid leukemia in hydroxyurea-treated sickle cell disease patient.

Hydroxyurea (HU) has been widely used in sickle cell disease. Its potential long-term risk for carcinogenesis or leukemogenic risk remains undefined. Here, we report a 26 y old African-American female with Sickle Cell Disease (SCD) who developed refractory/relapsed acute myeloid leukemia (AML) 6 months after 26 months of HU use. That patient's cytogenetics and molecular genetics analyses demonstrated a complex mutation profile with 5q deletion, trisomy 8, and P53 deletion (deletion of 17p13.1). P53 gene sequence studies revealed a multitude of somatic mutations that most suggest a treatment-related etiology. The above-mentioned data indicates that the patient may have developed acute myeloid leukemia with myelodysplasia-related changes (AML-MRC) as a direct result of HU exposure.

The Combined Effect of FGFR Inhibition and PD-1 Blockade Promotes Tumor-Intrinsic Induction of Antitumor Immunity.

The success of targeted or immune therapies is often hampered by the emergence of resistance and/or clinical benefit in only a subset of patients. We hypothesized that combining targeted therapy with immune modulation would show enhanced antitumor responses. Here, we explored the combination potential of erdafitinib, a fibroblast growth factor receptor (FGFR) inhibitor under clinical development, with PD-1 blockade in an autochthonous FGFR2K660N/p53mut lung cancer mouse model. Erdafitinib monotherapy treatment resulted in substantial tumor control but no significant survival benefit. Although anti-PD-1 alone was ineffective, the erdafitinib and anti-PD-1 combination induced significant tumor regression and improved survival. For both erdafitinib monotherapy and combination treatments, tumor control was accompanied by tumor-intrinsic, FGFR pathway inhibition, increased T-cell infiltration, decreased regulatory T cells, and downregulation of PD-L1 expression on tumor cells. These effects were not observed in a KRASG12C-mutant genetically engineered mouse model, which is insensitive to FGFR inhibition, indicating that the immune changes mediated by erdafitinib may be initiated as a consequence of tumor cell killing. A decreased fraction of tumor-associated macrophages also occurred but only in combination-treated tumors. Treatment with erdafitinib decreased T-cell receptor (TCR) clonality, reflecting a broadening of the TCR repertoire induced by tumor cell death, whereas combination with anti-PD-1 led to increased TCR clonality, suggesting a more focused antitumor T-cell response. Our results showed that the combination of erdafitinib and anti-PD-1 drives expansion of T-cell clones and immunologic changes in the tumor microenvironment to support enhanced antitumor immunity and survival.