Search

Abstract Background Unstructured text in medical records, such as Electronic Health Records, contain an enormous amount of valuable information for research; however, it is difficult to extract and structure important information because of frequent typographical errors. Therefore, improving the quality of data with errors for text analysis is an essential task. To date, few prior studies have been conducted addressing this. Here, we propose a new methodology for extracting important information from unstructured medical texts by overcoming the typographical problem in surgical pathology records related to lung cancer. Methods We propose a typo correction model that considers context, based on the Masked Language Model, to solve the problem of typographical errors in real-world medical data. In addition, a word dictionary was used for the typo correction model based on PubMed abstracts. After refining the data through typo correction, fine tuning was performed on pre-trained BERT model. Next, deep learning-based Named Entity Recognition (NER) was performed. By solving the quality problem of medical data, we sought to improve the accuracy of information extraction in unstructured text data. Results We compared the performance of the proposed typo correction model based on contextual information with an existing SymSpell model. We confirmed that our proposed model outperformed the existing model in a typographical correction task. The F1-score of the model improved by approximately 5% and 9% when compared with the model without contextual information in the NCBI-disease and surgical pathology record datasets, respectively. In addition, the F1-score of NER after typo correction increased by 2% in the NCBI-disease dataset. There was a significant performance difference of approximately 25% between the before and after typo correction in the Surgical pathology record dataset. This confirmed that typos influenced the information extraction of the unstructured text. Conclusion We verified that typographical errors in unstructured text negatively affect the performance of natural language processing tasks. The proposed method of a typo correction model outperformed the existing SymSpell model. This study shows that the proposed model is robust and can be applied in real-world environments by focusing on the typos that cause difficulties in analyzing unstructured medical text.

Abstract Background Studies on the application of targeted therapies for patients with non‐small cell lung cancer (NSCLC) who harbor rare genetic mutations are ongoing. In the present study, we investigated the real‐world data of NSCLC patients who harbor rare mutations. Methods We retrospectively analyzed patients with advanced or metastatic nonsquamous NSCLC aged >20 years with confirmed rare mutations (BRAF, ROS1, MET, RET, HER2, FGFR, and NTRK) from January 2015 to September 2020 at nine tertiary hospitals. In addition, we validated the lung cancer PCR panel kit in patients with confirmed mutations by NGS. Results Among 118 patients included, 88 received platinum‐based chemotherapy as first‐line chemotherapy. The progression‐free survival of patients with BRAF, ERBB2, MET, RET, and ROS1 mutations was 10.9 months (95% confidence interval [CI]: 1.3–20.5), 5.3 months (95% CI: 3.0–7.5), 7.2 months (95% CI: 3.6–10.9), 11.4 months (95% CI: 9.2–13.6), and 10.0 months (95% CI: 3.7–16.4) respectively (p = 0.041). The median overall survival (OS) was not reached in patients with ROS1 mutations; however, in BRAF, ERBB2, MET, and RET mutant patients, median OS was 14.1 months (95% CI: 10.1–14.1), 34.5 months (95% CI: 13.2–36.9), 22.7 months (95% CI: 1.7–24.0), and 29.8 months (95% CI: 28.9–61.3), respectively (p = 0.006). Of the 27 tissue samples, 26 (96.3%) showed the same PCR panel kit result with NGS. Conclusions First‐line platinum‐based chemotherapy showed durable benefit in patients with advanced or metastatic nonsquamous NSCLC harboring rare genetic mutation other than EGFR or ALK.

Abstract Objectives The optimal sequence for the administration of epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (TKIs) for treating non‐small cell lung cancer (NSCLC) is still unclear. This study aimed to evaluate the efficacy of sequential afatinib and osimertinib treatment in patients with NSCLC harboring EGFR mutations. Materials and methods Electronic records of patients with EGFR‐mutated NSCLC, who were administered afatinib and osimertinib (group A) or other chemotherapy (group B) between October 2014 and 2019, across 16 hospitals in South Korea were reviewed. The primary outcome, time on treatment (TOT), secondary outcome, and overall survival (OS) were estimated using the Kaplan–Meier method and log‐rank test. Multivariate analyses were performed using the Cox proportional hazards model. Results Of the 737 patients who received frontline afatinib treatment, 324 with complete records were selected (group A: 126, group B: 198). All patients in group A were T790M positive after afatinib, while patients in group B were all negative or unknown. The median TOT was 35.4 months (95% confidence interval [CI]: 27.7−45.6) in group A and 20.8 months (95% CI: 19.4−24.0) in group B. The median TOT with afatinib was 13.0 months (95% CI: 12.0−13.9) overall and 15.7 months (95% CI: 13.9−17.3) in group A. The 2‐ and 3‐year survival rates were 86.0 and 69.3% in group A and 75.9 and 55.3% in group B, respectively. Conclusion Sequential afatinib and osimertinib treatment resulted in better survival rates than treatment with afatinib followed by other chemotherapies.

Prediction models for major nutrients of rice were built using near-infrared (NIR) spectral data based on the artificial neural network (ANN). Scientific interpretation of the weight values was proposed and performed to understand the wavenumbers contributing to the prediction of nutrients. NIR spectra were acquired from 110 rice samples. Carbohydrate and moisture contents were predicted with values for the determination coefficient, relative root mean square error, range error ratio, and residual prediction deviation of 0.98, 0.11 %, 44, and 7.3, and 0.97, 0.80 %, 27, and 5.8, respectively. The results agreed well with ones reported in the previous studies and acquired by the conventional partial least squares (PLS)-variable importance in projection method. This study demonstrates that the combination of NIR and ANN is a powerful and accurate tool to monitor nutrients of rice and scientific interpretation of weights can be performed to overcome black box nature of the ANN.

This study aimed to investigate the feasibility of blood-based biomarkers, including blood tumor mutation burden (bTMB), to predict atezolizumab efficacy in relapsed and advanced non-small cell lung cancer (NSCLC). Stage IV NSCLC patients who had previously received platinum-doublet chemotherapy were recruited and received 1200 mg of atezolizumab every three weeks. Blood was collected to obtain plasma cell-free DNA (cfDNA) before the first cycle (C0) and at the fourth cycle (C4). bTMB was measured by CT-ULTRA in patients with cfDNA over 10 ng. The objective response rate (ORR) of the enrolled 100 patients was 10%, and there was no difference in ORR according to bTMB (cutoff: 11.5 muts/Mb) at C0 (high bTMB: 8.1% vs. low bTMB: 11.1%). However, the C4/C0 bTMB ratio was significantly lower in the durable clinical benefit (DCB) patients. The cfDNA concentration at C0, the C4/C0 ratio of the cfDNA concentration, the highest variant allele frequency (hVAF), and the VAF standard deviation (VAFSD) were significantly lower in the DCB patients. In the multivariate analysis, a high cfDNA concentration at C0 (cutoff: 8.6 ng/mL) and a C4/C0 bTMB ratio greater than 1 were significantly associated with progression-free survival. These results suggest that baseline levels and dynamic changes of blood-based biomarkers (bTMB, cfDNA concentration, and VAFSD) could predict atezolizumab efficacy in previously treated NSCLC patients.

Abstract Background EGFR genotyping in pulmonary adenocarcinoma patients who develop pleural effusions is mostly performed using cytology or cell block slides with low sensitivity. Liquid biopsy using the supernatant of pleural effusions may be more effective because they contain many components released by cancer cells. Extracellular vesicles (EVs) are known to carry oncogenic double-stranded DNA that is considered a notable biomarker. Here, we investigate the efficiency of liquid biopsy using cell-free DNA (cfDNA) and extracellular vesicle-derived DNA (EV-derived DNA) from the supernatant of pleural effusions for EGFR genotyping in patients with pulmonary adenocarcinoma. Methods Fifty pleural effusion samples from patients with pulmonary adenocarcinoma were evaluated. The supernatant, after removing the cell pellet by centrifugation, was used for liquid biopsy, and EVs were isolated from the pleural effusion by ultracentrifugation. EV-derived DNA and cfDNA were extracted separately, and EGFR genotyping was performed by the PNA clamping method. Results Among 32 patients who were EGFR-tyrosine kinase inhibitor (TKI) naïve with a known tissue EGFR genotype, liquid biopsy using EV-derived DNA from the pleural effusion supernatant showed 100% matching results with tissue EGFR genotyping in 19 EGFR mutant cases and detected three additional EGFR mutations in patients with wild-type (WT) tissue. Liquid biopsy using cfDNA from pleural effusion supernatants missed two cases of tissue-based EGFR mutations and found two additional EGFR mutation cases. In 18 patients who acquired resistance to EGFR-TKI, EGFR genotyping using EV-derived DNA from the pleural effusion supernatant detected the T790 M mutation in 13 of 18 (72.2%) patients, and this mutation was detected in 11 (61.1%) patients using cfDNA. By contrast, only three patients were found to present the T790 M mutation when using cell block or cytology slides. Conclusions Liquid biopsy using the supernatant of pleural effusions showed significantly improved results for EGFR genotyping compared to those using conventional cell block or cytology samples. Liquid biopsy using EV-derived DNA is promising for EGFR genotyping, including T790 M detection in pulmonary adenocarcinoma patients who develop pleural effusions.

Abstract Non-small cell lung cancer (NSCLC) is histologically classified into lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LSCC). However, some tumors are histologically ambiguous and other pathophysiological features or microenvironmental factors may be more prominent. Here we report integrative multiomics analyses using data for 229 patients from a Korean NSCLC cohort and 462 patients from previous multiomics studies. Histological examination reveals five molecular subtypes, one of which is a NSCLC subtype with PI3K-Akt pathway upregulation, showing a high proportion of metastasis and poor survival outcomes regardless of any specific NSCLC histology. Proliferative subtypes are present in LUAD and LSCC, which show strong associations with whole genome doubling (WGD) events. Comprehensive characterization of the immune microenvironment reveals various immune cell compositions and neoantigen loads across molecular subtypes, which predicting different prognoses. Immunological subtypes exhibit a hot tumor-enriched state and a higher efficacy of adjuvant therapy.

Protein kinase CK2 has diverse functions promoting and maintaining cancer phenotypes. We investigated the effect of CK2 inhibition in lung cancer cells with T790M-mediated resistance to the EGFR-TK inhibitor. Resistant sublines of PC-9 to gefitinib (PC-9/GR) and erlotinib (PC-9/ER) were established by previous study, and T790M secondary mutation was found in both resistant sublines. A decrease of EGFR by siRNA treatment effectively controlled the growth of resistant cells, thus suggesting that they still have EGFR-dependency. CX-4945, a potent and selective CK2 inhibitor, induced autophagy in PC-9/GR and PC-9/ER, and which was supported by the induction of autophagic vacuoles and microtubule-associated protein 1 light chain 3 (LC3) expression, and the increase of punctate fluorescent signals in resistant cells pre-transfected with green fluorescent protein (GFP)-tagged LC3. However, the withdrawal of CX-4945 led to the recovery of cancer cells with autophagy. We found that the induction of autophagy by CX-4945 in both resistant cells was CK2 dependent by using small interfering RNA against CK2. The treatment with CX-4945 alone induced a minimal growth inhibition in resistant cells. However, combined treatment of CX-4945 and EGFR-TKI effectively inhibited cancer-cell proliferation and induced apoptosis. CX-4945 increased the translocation of EGFR from the cell surface into the autophagosome, subsequently leading to the decrease of EGFR while inhibition of autophagy by 3MA or Atg7-targeted siRNA pretreatment reduced the decrease of EGFR by CX-4945. Accordingly, apoptosis by a combination of CX-4945 and EGFR-TKI was suppressed by 3MA or Atg7-targeted siRNA pretreatment, thus suggesting that autophagosome-mediated EGFR down-regulation would have an important role regarding apoptotic cell death by EGFR-TKI. Combined treatment of the CK2 inhibitor and EGFR-TKI may be a promising strategy for overcoming T790M-mediated resistance.

BACKGROUND: Epidemiologic data regarding tuberculin skin test (TST) responses are an important basis for TB control strategies. This study analyzed TST responses in Korea, which experienced a rapid change in BCG vaccination status. METHODS: TST responses in young adults were examined over 5 years. Participants with active TB lesions were excluded. RESULTS: A total of 5,552 participants were enrolled with median age of 21 years. When an induration diameter ≥10 mm was used as the criterion for a positive test, TST positivity fell (from 28.0% in 2005 to 15.3% in 2009); however, they remained steady when the criterion was ≥15-20 mm. A positive TST was associated with a personal or family of TB, the presence of a Bacille Calmette-Guérin (BCG) scar, and age (odds ratio [95% confidence interval] = 4.03 [2.61-6.22], 2.91 [1.80-4.71], 1.50 [1.31-1.72], and 1.15 [1.09-1.20], respectively). Among these factors, the decrease of participants with BCG scars was the most prominent change, which appeared to be associated with the change of TST positivity rate. CONCLUSION: Overall, the rate of TST positivity in Korea decreased. However, this trend seems associated with the change of BCG vaccination strategy rather than successful control of LTBI. This study showed that change in BCG vaccination strategy can have great impact on TB epidemiologic survey based on TST.

Abstract Purpose This study aimed to investigate the differences in characteristics, clinical stages, treatment modalities, and survival outcomes in patients with non‐small‐cell lung cancer (NSCLC) based on sex differences using Korean nationwide registry data. Methods We analyzed the data of 8650 patients diagnosed with NSCLC between 2014 and 2017, obtained from the Korean Association for Lung Cancer Registry (KALC‐R). The Cox proportional hazard model was used to define the differences in survival based on sex. Propensity score matching was used to adjust for differences between men and women. Results Of a total of 10 943 patients, 8650 (79.1%) were diagnosed with NSCLC, of whom 68.7% were men and 31.3% were women. For NSCLC, the median age was higher (69.0 vs. 67.0, p

Abstract Background Extracting metastatic information from previous radiologic-text reports is important, however, laborious annotations have limited the usability of these texts. We developed a deep-learning model for extracting primary lung cancer sites and metastatic lymph nodes and distant metastasis information from PET-CT reports for determining lung cancer stages. Methods PET-CT reports, fully written in English, were acquired from two cohorts of patients with lung cancer who were diagnosed at a tertiary hospital between January 2004 and March 2020. One cohort of 20,466 PET-CT reports was used for training and the validation set, and the other cohort of 4190 PET-CT reports was used for an additional-test set. A pre-processing model (Lung Cancer Spell Checker) was applied to correct the typographical errors, and pseudo-labelling was used for training the model. The deep-learning model was constructed using the Convolutional-Recurrent Neural Network. The performance metrics for the prediction model were accuracy, precision, sensitivity, micro-AUROC, and AUPRC. Results For the extraction of primary lung cancer location, the model showed a micro-AUROC of 0.913 and 0.946 in the validation set and the additional-test set, respectively. For metastatic lymph nodes, the model showed a sensitivity of 0.827 and a specificity of 0.960. In predicting distant metastasis, the model showed a micro-AUROC of 0.944 and 0.950 in the validation and the additional-test set, respectively. Conclusion Our deep-learning method could be used for extracting lung cancer stage information from PET-CT reports and may facilitate lung cancer studies by alleviating laborious annotation by clinicians.

Abstract Background Very few studies have compared the effects and side effects of vancomycin and teicoplanin in patients with methicillin-resistant Staphylococcus aureus pneumonia. This study aimed to compare the efficacy and safety of vancomycin and teicoplanin in patients with methicillin-resistant Staphylococcus aureus pneumonia. Methods This study examined 116 patients with methicillin-resistant Staphylococcus aureus pneumonia who met the inclusion criteria and were treated with either vancomycin (n = 54) or teicoplanin (n = 62). The primary (i.e., clinical failure during treatment) and secondary outcomes (i.e., mortality rates, discontinuation of study drugs due to treatment failure, side effects, and clinical cure) were evaluated. Results The vancomycin group presented lower clinical failure rates (25.9% vs. 61.3%, p

Abstract Background A previous trial showed that autologous ex‐vivo expanded NK cell (SNK01) treatment combined with pembrolizumab showed better efficacy than pembrolizumab monotherapy in advanced non‐small cell lung cancer (NSCLC). This study was a 2‐year follow‐up of that previous study to determine the long‐term efficacy of the combination treatment. Methods This trial included 20 patients with advanced NSCLC with a PD‐L1 tumor proportion score of 1% or greater who failed prior to front‐line platinum‐based therapy. The patients received pembrolizumab with low‐dose SNK01 (2 × 109 cells/dose) or high‐dose SNK01 (4 × 109 cells/dose), or pembrolizumab monotherapy. The primary study endpoint was overall survival (OS), and the secondary endpoint was progression‐free survival (PFS). Results Two patients were excluded following serious adverse events. Among the 11 patients who died, five were from the NK groups (41.6%, n = 5/12), and six received pembrolizumab monotherapy (100%, n = 6/6). The estimated 2‐year survival rate was 58.3% versus 16.7% (pembrolizumab plus SNK01 vs. pembrolizumab monotherapy). The hazard ratio of pembrolizumab plus SNK01 compared with pembrolizumab monotherapy was 0.32 (95% CI: 0.1, 1.08, p‐value: 0.066). Although the median PFS was significantly higher in the pembrolizumab plus SNK01 group than in the pembrolizumab alone group, OS and PFS did not differ statistically between patients who received low doses of NK cells and those who received high doses of NK cells. Conclusions Autologous NK cells can enhance the long‐term OS and PFS for NSCLC. A larger study is needed to confirm this result. Clinical Research Information Service number: KCT0003463.

Background: Hominis placenta pharmacopuncture is widely used for climacteric symptoms. This study examined the efficacy and safety of pharmacopuncture with PLC (the extract of Hominis placenta) on hot flashes for perimenopausal and postmenopausal women. Methods: This study was a randomized placebo-controlled single-blind trial, which recruited 128 perimenopausal and postmenopausal women, randomly assigned to receive pharmacopuncture with PLC or normal saline (NS) for eight weeks. The primary outcome was the mean changes in the hot flash score (HFS) and the secondary outcomes were the mean changes in the Menopause Rating Scale (MRS), follicle-stimulating hormone (FSH) levels, and estradiol (E2) levels from baseline to eight weeks. Missing values were imputed using the last-observation-carried-forward method. Results: After treatment (week 9), the HFS decreased significantly in both groups (p = 0.000). The residual HFS was 47.09 ± 41.39% and 56.45 ± 44.92 % in the PLC and control groups, respectively (p = 0.262). One month after the treatment (week 13), the score of the PLC group was reduced, but the score increased in the control group (p = 0.077). There were no statistically significant differences in the mean changes in MRS, FSH, and E2 between the two groups. No serious adverse events related to this trial were noted. Conclusion: In this study, Hominis placenta extract pharmacopuncture did not differ significantly from NS in reducing the hot flash score. While this therapy appears safe, the potential for long-term effect of PLC extract needs to be examined in a large randomized controlled trial with appropriate controls. Trial registration: Clinical Research Information Service, Republic of Korea, KCT0 0 03533.

Purpose: This study aimed to assess the real-world clinical outcomes of consolidative durvalumab in patients with unresectable locally advanced non-small cell lung cancer (LA-NSCLC) and to explore the role of radiotherapy in the era of immunotherapy. Materials and Methods: This retrospective study assessed 171 patients with unresectable LA-NSCLC who underwent concurrent chemoradiotherapy (CCRT) with or without consolidative durvalumab at Asan Medical Center between May 2018 and May 2021. Primary outcomes included freedom from locoregional failure (FFLRF), distant metastasis-free survival (DMFS), progression-free survival (PFS), and overall survival (OS). Results: Durvalumab following CCRT demonstrated a prolonged median PFS of 20.9 months (p=0.048) and a 3-year FFLRF rate of 57.3% (p=0.008), compared to 13.7 months and 38.8%, respectively, with CCRT alone. Furthermore, the incidence of in-field recurrence was significantly greater in the CCRT-alone group compared to the durvalumab group (26.8% vs. 12.4%, p=0.027). While median OS was not reached with durvalumab, it was 35.4 months in patients receiving CCRT alone (p=0.010). Patients positive for programmed cell death ligand 1 (PD-L1) expression showed notably better outcomes, including FFLRF, DMFS, PFS, and OS. Adherence to PACIFIC trial eligibility criteria identified 100 patients (58.5%) as ineligible. The use of durvalumab demonstrated better survival regardless of eligibility criteria. Conclusion The use of durvalumab consolidation following CCRT significantly enhanced locoregional control and OS in patients with unresectable LA-NSCLC, especially in those with PD-L1-positive tumors, thereby validating the role of durvalumab in standard care. [ABSTRACT FROM AUTHOR]

Purpose: This study aimed to report the results from an early-phase study of rivoceranib, an oral tyrosine kinase inhibitor highly selective for vascular endothelial growth factor receptor 2, in patients with advanced solid tumors. Materials and Methods: In this open-label, single-arm, dose-escalating, multicenter three-part phase 1/2a trial, patients had advanced solid tumors refractory to conventional therapy. Part 1 evaluated the safety and pharmacokinetics of five ascending oncedaily doses of rivoceranib from 81 mg to 685 mg. Part 2 evaluated the safety and antitumor activity of once-daily rivoceranib 685 mg. Part 3 was conducted later, due to lack of maximum tolerated dose determination in part 1, to evaluate the safety and preliminary efficacy of once-daily rivoceranib 805 mg in patients with unresectable or advanced gastric cancer. Results: A total of 61 patients were enrolled in parts 1 (n=25), 2 (n=30), and 3 (n=6). In parts 1 and 2, patients were white (45.5%) or Asian (54.5%), and 65.6% were male. The most common grade ≥ 3 adverse events were hypertension (32.7%), hyponatremia (10.9%), and hypophosphatemia (10.9%). The objective response rate (ORR) was 15.2%. In part 3, dose-limiting toxicities occurred in two out of six patients: grade 3 febrile neutropenia decreased appetite, and fatigue. The ORR was 33%. Conclusion The recommended phase 2 dose of rivoceranib was determined to be 685 mg once daily, which showed adequate efficacy with a manageable safety profile (NCT01497704 and NCT02711969). [ABSTRACT FROM AUTHOR]

Lung cancer remains the most common cause of cancer-related deaths worldwide. Although there are many possible treatments, including targeted therapies such as epidermal growth factor receptor tyrosine kinase inhibitors and anaplastic lymphoma kinase inhibitors, new therapeutic strategies are needed to improve clinical outcomes. Immunotherapy through the use of immune checkpoint inhibitors has provided one of the most important breakthroughs in the management of solid tumors, including lung cancers, and has shown promising results in numerous clinical trials. This review will present the current status of immunotherapy for lung cancer and future perspectives on these treatments.

Electromagnetic navigation bronchoscopy (ENB) is an emerging technique used to evaluate peripheral lung lesions. The aim of this study was to determine the diagnostic yield, safety profile, and adequacy of specimens obtained using ENB for molecular testing. This single-center, prospective pilot study recruited patients with peripheral pulmonary nodules that were not suitable for biopsy via percutaneous transthoracic needle biopsy methods. The possibility of molecular testing, including epidermal growth factor receptor (EGFR), anaplastic lymphoma kinase (ALK), and programmed death ligand 1 (PD-L1), was identified with non-small cell lung cancer (NSCLC) tissue obtained using ENB. ENB-guided biopsy was performed on 30 pulmonary nodules in 30 patients. ENB-guided biopsy was successfully performed in 96.6% (29/30) of cases, but one case failed to approach the target lesion. The diagnostic accuracy of ENB-guided biopsy was 68.0% (17/25). Biopsy-related pneumothorax occurred in one patient and there was no major bleeding or deaths related to the procedure. Among 13 patients diagnosed with NSCLC, molecular testing was successfully performed in 92.3% (12/13). ENB-guided biopsy demonstrated acceptable accuracy and excellent sample adequacy, with a high possibility of achieving molecular testing and a good safety profile to evaluate peripheral pulmonary nodules, even when the percutaneous approach was difficult and/or dangerous.