ABSTRACT
Purpose@#We aimed to evaluate the time and cost of developing prompts using large language model (LLM), tailored to extract clinical factors in breast cancer patients and their accuracy. @*Materials and Methods@#We collected data from reports of surgical pathology and ultrasound from breast cancer patients who underwent radiotherapy from 2020 to 2022. We extracted the information using the Generative Pre-trained Transformer (GPT) for Sheets and Docs extension plugin and termed this the “LLM” method. The time and cost of developing the prompts with LLM methods were assessed and compared with those spent on collecting information with “full manual” and “LLM-assisted manual” methods. To assess accuracy, 340 patients were randomly selected, and the extracted information by LLM method were compared with those collected by “full manual” method. @*Results@#Data from 2,931 patients were collected. We developed 12 prompts for Extract function and 12 for Format function to extract and standardize the information. The overall accuracy was 87.7%. For lymphovascular invasion, it was 98.2%. Developing and processing the prompts took 3.5 hours and 15 minutes, respectively. Utilizing the ChatGPT application programming interface cost US $65.8 and when factoring in the estimated wage, the total cost was US $95.4. In an estimated comparison, “LLM-assisted manual” and “LLM” methods were time- and cost-efficient compared to the “full manual” method. @*Conclusion@#Developing and facilitating prompts for LLM to derive clinical factors was efficient to extract crucial information from huge medical records. This study demonstrated the potential of the application of natural language processing using LLM model in breast cancer patients. Prompts from the current study can be re-used for other research to collect clinical information.
ABSTRACT
Purpose@#This study aimed to develop a model for predicting pathologic extracapsular extension (ECE) and seminal vesicle invasion (SVI) while integrating magnetic resonance imaging-based T-staging (cTMRI, cT1c-cT3b). @*Materials and Methods@#A total of 1,915 who underwent radical prostatectomy between 2006-2016 met the inclusion/exclusion criteria. We performed a multivariate logistic regression analysis as well as Bayesian network (BN) modeling based on possible confounding factors. The BN model was internally validated using 5-fold validation. @*Results@#According to the multivariate logistic regression analysis, initial prostate-specific antigen (iPSA) (β=0.050, p < 0.001), percentage of positive biopsy cores (PPC) (β=0.033, p < 0.001), both lobe involvement on biopsy (β=0.359, p=0.009), Gleason score (β=0.358, p < 0.001), and cTMRI (β=0.259, p < 0.001) were significant factors for ECE. For SVI, iPSA (β=0.037, p < 0.001), PPC (β=0.024, p < 0.001), Gleason score (β=0.753, p < 0.001), and cTMRI (β=0.507, p < 0.001) showed statistical significance. BN models to predict ECE and SVI were also successfully established. The overall area under the receiver operating characteristic curve (AUC)/accuracy of the BN models were 0.76/73.0% and 0.88/89.6% for ECE and SVI, respectively. According to internal comparison between the BN model and Roach formula, BN model had improved AUC values for predicting ECE (0.76 vs. 0.74, p=0.060) and SVI (0.88 vs. 0.84, p < 0.001). @*Conclusion@#Two models to predict pathologic ECE and SVI integrating cTMRI were established and installed on a separate website for public access to guide radiation oncologists.
ABSTRACT
Purpose@#The value of the genomic profiling by targeted gene-sequencing on radiation therapy response prediction was evaluated through integrated analysis including clinical information. Radiation response prediction model was constructed based on the analyzed findings. @*Materials and Methods@#Patients who had the tumor sequenced using institutional cancer panel after informed consent and received radiotherapy for the measurable disease served as the target cohort. Patients with irradiated tumor locally controlled for more than 6 months after radiotherapy were defined as the durable local control (DLC) group, otherwise, non-durable local control (NDLC) group. Significant genomic factors and domain knowledge were used to develop the Bayesian Network model to predict radiotherapy response. @*Results@#Altogether, 88 patients were collected for analysis. Of those, 41 (43.6%) and 47 (54.4%) patients were classified as the NDLC and DLC group, respectively. Somatic mutations of NOTCH2 and BCL were enriched in the NDLC group, whereas, mutations of CHEK2, MSH2, and NOTCH1 were more frequently found in the DLC group. Altered DNA repair pathway was associated with better local failure–free survival (hazard ratio, 0.40; 95% confidence interval, 0.19 to 0.86; p=0.014). Smoking somatic signature was found more frequently in the DLC group. Area under the receiver operating characteristic curve of the Bayesian network model predicting probability of 6-month local control was 0.83. @*Conclusion@#Durable radiation response was associated with alterations of DNA repair pathway and smoking somatic signature. Bayesian network model could provide helpful insights for high precision radiotherapy. However, these findings should be verified in prospective cohort for further individualization.
ABSTRACT
Radiation therapy for patients with pN1mi or pN1 disease breast cancer undergoing mastectomy has been debated for a long time. Even in low metastatic burden in sentinel node biopsy, occult non-sentinel axillary nodal involvement can exist. Radiotherapy can sterilize axillary metastatic burden and seems to contribute a very low local recurrence rate in mastectomy patients with minimally involved lymph nodes. However, it should be considered that systemic therapy is evolving and the local recurrence difference between radiotherapy and no radiotherapy is relatively small. Regarding postmastectomy radiotherapy in patients pN1mi or pN1 cancer, published prospective clinical trial results should be considered; however, there are no such relevant results of clinical trials yet. Consideration of postmastectomy radiation therapy in pN1mi or pN1 patients should be based on identifying the high-risk group in terms of recurrence, stage, or tumor biology. When radiotherapy is determined, radiation oncologists should attempt individualized treatment approaches, such as irradiation field, and consider specific settings, such as neoadjuvant therapy. In this review, the role of radiotherapy in mastectomy patients with minimally involved lymph nodes and the relevant considerations are discussed.
ABSTRACT
Purpose@#Combination of radiotherapy and immune checkpoint blockade such as programmed death-1 (PD-1) or programmed death-ligand 1 (PD-L1) blockade is being actively tested in clinicaltrial. We aimed to identify a subset of patients that could potentially benefit from this strategyusing The Cancer Genome Atlas (TCGA) dataset for glioblastoma (GBM). @*Materials and Methods@#A total of 399 cases were clustered into radiosensitive versus radioresistant (RR) groupsbased on a radiosensitivity gene signature and were also stratified as PD-L1 high versusPD-L1 low groups by expression of CD274 mRNA. Differential and integrated analyses withexpression and methylation data were performed. CIBERSORT was used to enumerate theimmune repertoire that resulted from transcriptome profiles. @*Results@#We identified a subset of GBM, PD-L1-high-RR group which showed worse survival comparedto others. In PD-L1-high-RR, differentially expressed genes (DEG) were highly enriched forimmune response and mapped into activation of phosphoinositide 3-kinase–AKT andmitogen-activated protein kinase (MAPK) signaling pathways. Integration of DEG and differentiallymethylated region identified that the kinase MAP3K8-involved in T-cell receptor signalingwas upregulated and BAI1, a factor which inhibits angiogenesis, was silenced.CIBERSORT showed that a higher infiltration of the immune repertoire, which included M2macrophages and regulatory T cells. @*Conclusion@#Taken together, PD-L1-high-RR group could potentially benefit from radiotherapy combinedwith PD-1/PD-L1 blockade and angiogenesis inhibition.
ABSTRACT
PURPOSE: We evaluated failure pattern and treatment outcomes of observational approach on regional lymph node (LN) in cutaneous melanoma of extremities and sought to find clinico-pathologic factors related to LN metastases. MATERIAL AND METHODS: We retrospectively reviewed 73 patients with cutaneous melanoma of extremities between 2005 and 2016. If preoperative 18-F-fluorodeoxyglucose (FDG)-positron emission tomography/computed tomography (PET/CT) findings were non-specific for regional LNs, surgical resection of primary tumors with adequate margins was performed without sentinel lymph node biopsy (SLNB) and/or complete lymph node dissection (CLND), irrespective of tumor thickness or size. In patients with suspicious or positive findings on PET/CT or CT, SLNB followed by CLND or CLND was performed at the discretion of the surgeon. We defined LN dissection (LND) as SLNB and/or CLND. RESULTS: With a median follow-up of 38 months (range, 6 to 138 months), the dominant pattern of failure was regional failure (17 of total 23 events, 74%) in the observation group (n = 56). Pathologic LN metastases were significant factor for poor regional failure-free survival (hazard ration [HR] = 3.21; 95% confidence interval [CI], 1.03–10.33; p = 0.044) and overall survival (HR = 3.62; 95% CI, 1.02–12.94; p = 0.047) in multivariate analysis. In subgroup analysis for cN0 patients according to the preoperative PET/CT findings, LND group showed the better trend of LRFFS (log rank test, p = 0.192) and RFFS (p = 0.310), although which is not statistically significant. CONCLUSION: Observational approach on regional LNs on the basis of the PET/CT in patients with cutaneous melanoma of extremities showed the dominant regional failure pattern compared to upfront LND approach. To reveal regional lymph node status, SLND for cN0 patients may of importance in managing cutaneous melanoma patients.
Subject(s)
Humans , Extremities , Follow-Up Studies , Lymph Node Excision , Lymph Nodes , Melanoma , Multivariate Analysis , Neoplasm Metastasis , Positron Emission Tomography Computed Tomography , Retrospective Studies , Sentinel Lymph Node BiopsyABSTRACT
PURPOSE: We aimed to evaluate clinical outcomes including progression-free survival (PFS), overall survival (OS), partial response, and complete response in patients who underwent radiation therapy (RT) for mycosis fungoides (MF). Also, we sought to find prognostic factors for clinical outcomes. MATERIALS AND METHODS: Total 19 patients confirmed with MF between 1999–2015 were retrospectively reviewed. Clinical and treatment characteristics, clinical outcomes, and and toxicities were analyzed. RESULTS: Eleven patients were treated with total skin electron beam radiotherapy (TSEBT) and 8 patients with involved field radiation therapy (IFRT) with median dose of 30 Gy, respectively. The median time interval from diagnosis to RT was 2.6 months (range, 0.4 to 87.3 months). The overall response rate was 100%; 11 patients (57.9%) had a complete response and 8 patients (42.1%) a partial response. The presence of positive lymph node at the time of consultation of RT was associated with lower OS (p = 0.043). In multivariate analysis, PFS was significantly lower for patients with increased previous therapies experienced following RT (p = 0.019) and for patients showing PR during RT (p = 0.044). There were no reported grade 3 or more skin toxicities related with RT. CONCLUSION: Both IFRT and TSEBT are effective treatment for MF patients. Patients with short disease course before RT or complete response during RT are expected to have longer PFS. Positive lymph node status at the initiation of RT was associated woth poor OS, suggesting other treatment modalities such as low-dose RT for patients with low life-expectancy.