Standardized Classification of Lung Adenocarcinoma Subtypes and Improvement of Grading Assessment Through Deep Learning.

The histopathologic distinction of lung adenocarcinoma (LADC) subtypes is subject to high interobserver variability, which can compromise the optimal assessment of patient prognosis. Therefore, this study developed convolutional neural networks capable of distinguishing LADC subtypes and predicting disease-specific survival, according to the recently established LADC tumor grades. Consensus LADC histopathologic images were obtained from 17 expert pulmonary pathologists and one pathologist in training. Two deep learning models (AI-1 and AI-2) were trained to predict eight different LADC classes. Furthermore, the trained models were tested on an independent cohort of 133 patients. The models achieved high precision, recall, and F1 scores exceeding 0.90 for most of the LADC classes. Clear stratification of the three LADC grades was reached in predicting the disease-specific survival by the two models, with both Kaplan-Meier curves showing significance (P = 0.0017 and 0.0003). Moreover, both trained models showed high stability in the segmentation of each pair of predicted grades with low variation in the hazard ratio across 200 bootstrapped samples. These findings indicate that the trained convolutional neural networks improve the diagnostic accuracy of the pathologist and refine LADC grade assessment. Thus, the trained models are promising tools that may assist in the routine evaluation of LADC subtypes and grades in clinical practice.

Frequency of subclinical interstitial lung disease in COVID-19 autopsy cases: potential risk factors of severe pneumonia.

Risk factors of severe coronavirus disease 2019 (COVID-19) have been previously reported; however, histological risk factors have not been defined thus far. The aim of this study was to clarify subclinical hidden interstitial lung disease (ILD) as a risk factor of severe pneumonia associated with COVID-19. We carefully examined autopsied lungs and chest computed tomography scanning (CT) images from patients with COVID-19 for interstitial lesions and then analyzed their relationship with disease severity. Among the autopsy series, subclinical ILD was found in 13/27 cases (48%) in the COVID-19 group, and in contrast, 8/65 (12%) in the control autopsy group (p = 0.0006; Fisher's exact test). We reviewed CT images from the COVID-19 autopsy cases and verified that subclinical ILD was histologically detectable in the CT images. Then, we retrospectively examined CT images from another series of COVID-19 cases in the Yokohama, Japan area between February-August 2020 for interstitial lesions and analyzed the relationship to the severity of COVID-19 pneumonia. Interstitial lesion was more frequently found in the group with the moderate II/severe disease than in the moderate I/mild disease (severity was evaluated according to the COVID-19 severity classification system of the Ministry of Health, Labor, and Welfare [Japan]) (moderate II/severe, 11/15, 73.3% versus moderate I/mild, 108/245, 44.1%; Fisher exact test, p = 0.0333). In conclusion, it was suggested that subclinical ILD could be an important risk factor for severe COVID-19 pneumonia. A benefit of these findings could be the development of a risk assessment system using high resolution CT images for fatal COVID-19 pneumonia.

Impact of bronchoalveolar lavage lymphocytosis on the effects of anti-inflammatory therapy in idiopathic non-specific interstitial pneumonia, idiopathic pleuroparenchymal fibroelastosis, and unclassifiable idiopathic interstitial pneumonia.

BACKGROUND: Idiopathic non-specific interstitial pneumonia (iNSIP), idiopathic pleuroparenchymal fibroelastosis (iPPFE), and unclassifiable idiopathic interstitial pneumonia (IIP) are IIPs with chronic fibrotic phenotypes, and unlike idiopathic pulmonary fibrosis, they have often been treated with anti-inflammatory drugs, including corticosteroids and immunosuppressants. However, the impact of bronchoalveolar lavage (BAL) lymphocytosis on the effects of anti-inflammatory therapy has never been evaluated. This study aimed to elucidate whether BAL lymphocytosis can be used to predict the efficacy of anti-inflammatory drugs for iNSIP, iPPFE, and unclassifiable IIP. METHODS: Japanese patients diagnosed with iNSIP, iPPFE, and unclassifiable IIP by multidisciplinary discussion were identified using the nationwide registry. Eligible patients were stratified into four groups with and without BAL lymphocytosis and anti-inflammatory therapy to compare overall survival (OS) and changes in lung function. BAL lymphocytosis was defined as a lymphocyte differential count > 15%, and the cut-off was corroborated by survival classification and regression tree analysis. RESULTS: Overall, 186 patients (37 iNSIP, 16 iPPFE, and 133 unclassifiable IIP) were analyzed. Limited to patients treated with anti-inflammatory drugs (n = 123), patients with BAL lymphocytosis had a better prognosis [hazard ratio (HR), 0.26; 95% confidence interval (CI), 0.11-0.63; P = 0.003], higher slope of forced vital capacity (FVC) % predicted for 2 years, and longer OS (log-rank test, P = 0.012) than those without BAL lymphocytosis. On multivariate analysis, BAL lymphocytosis (HR 0.31; 95% CI 0.13-0.75; P = 0.009) was a prognostic factor for OS, along with age and FVC % predicted. Conversely, for patients managed without anti-inflammatory therapy (n = 63), the presence or absence of BAL lymphocytosis had no prognostic value. CONCLUSIONS: BAL lymphocytosis is associated with good outcomes in patients treated with anti-inflammatory drugs, but has no prognostic value when anti-inflammatory drugs are not used. BAL lymphocytosis may provide a predictive biomarker for identifying patients with iNSIP, iPPFE and unclassifiable IIP who are likely to benefit from anti-inflammatory drugs.

Probable usual interstitial pneumonia pattern on chest CT: is it sufficient for a diagnosis of idiopathic pulmonary fibrosis?

Recent studies have suggested that in patients with an idiopathic interstitial pneumonia (IIP), a probable usual interstitial pneumonia (UIP) pattern on chest computed tomography (CT) is sufficient to diagnose idiopathic pulmonary fibrosis (IPF) without histopathology.We retrospectively compared the prognosis and time to first acute exacerbation (AE) in IIP patients with a UIP and a probable UIP pattern on initial chest CT.One hundred and sixty IIP patients with a UIP pattern and 242 with a probable UIP pattern were identified. Probable UIP pattern was independently associated with longer survival time (adjusted hazard ratio 0.713, 95% CI 0.536-0.950; p=0.021) and time to first AE (adjusted hazard ratio 0.580, 95% CI 0.389-0.866; p=0.008). In subjects with a probable UIP pattern who underwent surgical lung biopsy, the probability of a histopathological UIP pattern was 83%. After multidisciplinary discussion and the inclusion of longitudinal behaviour, a diagnosis of IPF was made in 66% of cases. In IPF patients, survival time and time to first AE were not associated with CT pattern. Among subjects with a probable UIP pattern, compared to non-IPF patients, survival time and time to first AE were shorter in IPF patients.In conclusion, IIP patients with a probable UIP pattern on initial chest CT had a better prognosis and longer time to first AE than those with a UIP pattern. However, when baseline data and longitudinal behaviour provided a final diagnosis of IPF, CT pattern was not associated with these outcomes. This suggests diagnostic heterogeneity among patients with a probable UIP pattern.

Acute exacerbations of fibrotic interstitial lung diseases.

BACKGROUND AND OBJECTIVE: Acute exacerbation (AE) is a severe complication of idiopathic pulmonary fibrosis (AE-IPF). In 2016, an international working group revised its definition and diagnostic criteria; however, few studies have assessed the frequency and prognosis of AE in patients with other fibrotic interstitial lung diseases (FILD). METHODS: We used data from 1019 consecutive interstitial lung disease (ILD) patients initially evaluated between January 2008 and July 2015. All subject diagnoses were made by multidisciplinary discussion in December 2018. ILD was categorized as IPF (n = 462) and other FILD which included non-specific interstitial pneumonia (n = 22), chronic hypersensitivity pneumonitis (n = 29), connective tissue disease-associated ILD (n = 205) and unclassifiable ILD (n = 209). Using the 2016 definition of AE-IPF, we identified all subjects with an AE. RESULTS: During the observational period, 193 patients experienced a first AE (AE-FILD n = 69, AE-IPF n = 124). The time to first AE was significantly longer in FILD than IPF (log-rank test, P < 0.001). After adjusting for potentially influential confounders, FILD remained a significant predictor of longer time to first AE compared with IPF (hazard ratio: 0.453; 95% CI: 0.317-0.647, P = 0.006). In a multivariate Cox proportional analysis, baseline disease severity was closely associated with the incidence of AE-ILD. Even after adjustment for other clinical variables, AE had a negative impact on overall survival. AE-FILD and AE-IPF showed similar poor short-term outcomes. CONCLUSION: All forms of ILD are at risk of AE and have a similar outcome to AE-IPF.

Nationwide cloud-based integrated database of idiopathic interstitial pneumonias for multidisciplinary discussion.

Multidisciplinary discussion (MDD) requiring close communication between specialists (clinicians, radiologists and pathologists) is the gold standard for the diagnosis of idiopathic interstitial pneumonias (IIPs). However, MDD by specialists is not always feasible because they are often separated by time and location. An online database would facilitate data sharing and MDD. Our aims were to develop a nationwide cloud-based integrated database containing clinical, radiological and pathological data of patients with IIPs along with a web-based MDD system, and to validate the diagnostic utility of web-based MDD in IIPs.Clinical data, high-resolution computed tomography images and lung biopsy slides from patients with IIPs were digitised and uploaded to separate servers to develop a cloud-based integrated database. Web-based MDD was performed using the database and video-conferencing to reach a diagnosis.Clinical, radiological and pathological data of 524 patients in 39 institutions were collected, uploaded and incorporated into the cloud-based integrated database. Subsequently, web-based MDDs with a pulmonologist, radiologist and pathologist using the database and video-conferencing were successfully performed for the 465 cases with adequate data. Overall, the web-based MDD changed the institutional diagnosis in 219 cases (47%). Notably, the MDD diagnosis yielded better prognostic separation among the IIPs than did the institutional diagnosis.This is the first study of developing a nationwide cloud-based integrated database containing clinical, radiological and pathological data for web-based MDD in patients with IIPs. The database and the web-based MDD system that we built made MDD more feasible in practice, potentially increasing accurate diagnosis of IIPs.

Ki-67 is a strong prognostic marker of non-small cell lung cancer when tissue heterogeneity is considered.

BACKGROUND: Ki-67 expression is a well-established prognostic marker in various cancers. However, Ki-67 expression is also known as being heterogeneous. We investigated the prognostic significance of Ki-67 from the view of staining heterogeneity by the technique of Spiral Array. METHODS: 100 cases of resected lung cancer from Toyama university hospital archive were collected. Spiral Array blocks were generated out of 100 cases using 100 µm thick paraffin sections. Four µm thick sections of the Array block were stained for Ki-67. Staining results in each reel were scored for areas with lowest (LS), highest (HS), and average (AS) expression, exclusively in the cancer cells. Heterogeneity score (HeS) was designed as the difference between HS and LS. The scores were divided into four grades (0-3). Clinical information was collected, and the prognostic significance of Ki-67 was analyzed. RESULTS: Pathological stage was available for 91 patients (43 stage IA, 22 stage IB, 2 stage IIA, 9 stage IIB, 13 stage IIIA, 1 stage IIIB, and 1 stage IV). The HS of Ki-67 score in non-small cell lung cancer was 3 in 17 cases, 2 in 27 cases, 1 in 28 cases, 0 in 21 cases, and 4 reels were lost. 78 cases had clinical follow up. 74 cases had all the information available and were analyzed for correlation between Ki-67 expression and survival. Cases with score 2 and 3 of HS and HeS showed significant poorer prognosis (both P < 0.001), whereas LS or AS did not show significance. The results were identical when analyzing adenocarcinoma and squamous cell carcinoma, separately. Cox multivariate analysis of Ki-67 showed that HS was an independent risk factor affecting overall survival. CONCLUSIONS: Ki-67 is a strong prognostic marker for non-small cell lung cancer when the degree of highest staining frequency or heterogeneity is considered.

Heterotopic salivary duct inclusions in Warthin tumor - A cryptic histological finding involved in its pathogenesis.

Warthin tumor (WT) is the second most common benign parotid gland tumor after pleomorphic adenoma. WT is characterized by cystic and papillary proliferation of a two-layered oncocytic epithelium supported by lymphoid tissue. Heterotopic salivary duct inclusions (SDIs) are frequently observed in lymph nodes (LNs) of WT (SDI/LNs), and are thought to be the origin of WT. If this is true, SDIs should also persist in the lymphoid tissue of WT itself (SDI/WT), as a missing link between SDIs and WTs, but studies of this issue are limited. From 2008-2023, 138 WT cases were surgically excised at our hospital. SDI/LNs and SDI/WTs were histologically examined. Of 100 WT cases with LNs, SDI/LNs were observed in 67 cases (67 %). SDI/WTs were detected in 114 of 138 cases (82.6 %), including 107 of 127 smokers (84.3 %) and 7 of 11 never-smokers (63.6 %). SDI/WTs were located mainly in the subcapsular lymphoid tissue and often surrounded by a fibrous coat resembling salivary excretory ducts. This study revealed a high incidence of SDIs in WT itself, strongly supporting the theory that WT develops from heterotopic salivary ducts.

Variability of radiological and clinical features in cases with usual interstitial pneumonia without honeycombing.

BACKGROUND: Usual interstitial pneumonia (UIP) cases without honeycombing (possible UIP) included various CT features and was often difficult to diagnose. PURPOSE: This study aimed to classify the cases with possible UIP on CT features using cluster analysis and evaluate the features of subsets of participants and the correlation of prognosis. MATERIALS AND METHODS: The study included 85 patients with possible UIP in the 2011 idiopathic pulmonary fibrosis (IPF) guideline with radiological diagnosis. All cases underwent surgical biopsies and were diagnosed by multidisciplinary discussion (MDD) from the nationwide registry in Japan. The readers evaluated pulmonary opacity, nodules, cysts, and predominant distribution which were reclassified by IPF guidelines in 2018. Additionally, cases were classified into four groups by cluster analysis based on CT findings. The differences in survival among IPF classification and the clusters were evaluated. RESULTS: Cases were diagnosed as IPF (n = 55), NSIP (n = 4), unclassifiable (n = 23), and others (n = 3) by MDD. Cluster analysis revealed 4 clusters by CT features (n = 47, 16, 19 and 3, respectively). Cluster 1 had fewer lesions overall. Cluster 2 have many pure ground-glass opacities and ground-glass opacities with reticulation. Cluster 3 had many reticular opacities and nodules with few lower predominant distributions. Cluster 4 was characterized by peribronchovascular consolidation.The mean survival time of cluster 1 (4518 days) was significantly better than cluster 2, 3, and 4 (1843, 2196, and 1814 days, respectively) (p = 0.03). CONCLUSION: In conclusion, UIP without honeycombing included various CT patterns and MDD diagnoses. Significangly differences in prognosis were observed among clusters classified by CT findings.

Machine-Learning-Based Classification Model to Address Diagnostic Challenges in Transbronchial Lung Biopsy.

BACKGROUND: When obtaining specimens from pulmonary nodules in TBLB, distinguishing between benign samples and mis-sampling from a tumor presents a challenge. Our objective is to develop a machine-learning-based classifier for TBLB specimens. METHODS: Three pathologists assessed six pathological findings, including interface bronchitis/bronchiolitis (IB/B), plasma cell infiltration (PLC), eosinophil infiltration (Eo), lymphoid aggregation (Ly), fibroelastosis (FE), and organizing pneumonia (OP), as potential histologic markers to distinguish between benign and malignant conditions. A total of 251 TBLB cases with defined benign and malignant outcomes based on clinical follow-up were collected and a gradient-boosted decision-tree-based machine learning model (XGBoost) was trained and tested on randomly split training and test sets. RESULTS: Five pathological changes showed independent, mild-to-moderate associations (AUC ranging from 0.58 to 0.75) with benign conditions, with IB/B being the strongest predictor. On the other hand, FE emerged to be the sole indicator of malignant conditions with a mild association (AUC = 0.66). Our model was trained on 200 cases and tested on 51 cases, achieving an AUC of 0.78 for the binary classification of benign vs. malignant on the test set. CONCLUSION: The machine-learning model developed has the potential to distinguish between benign and malignant conditions in TBLB samples excluding the presence or absence of tumor cells, thereby improving diagnostic accuracy and reducing the burden of repeated sampling procedures for patients.

Minimally invasive elective gastrectomy after preoperative chemotherapy in a patient with frailty who presented with locally far advanced-stage gastric cancer: a case report.

BACKGROUND: Herein, we report a case of gastric antrum cancer with multiple invasions to other organs that was completely cured with laparoscopic distal gastrectomy after preoperative chemotherapy in a patient with poor general condition. CASE PRESENTATION: An 80-year-old male patient was diagnosed with anemia during follow-up for cerebral lacunar infarction at another hospital. He was diagnosed with advanced-stage gastric antrum cancer and was referred to our hospital. On esophagogastroduodenoscopy, type 2 advanced-stage gastric cancer was detected at the greater curvature of the antrum, and the biopsy results revealed tubular adenocarcinoma. Contrast-enhanced computed tomography scan revealed multiple invasions to other organs, thick gastric wall with contrast effect, and superior mesenteric vein tumor thrombus. However, there was no evidence of distant metastasis on positron emission tomography/computed tomography scan. The clinical diagnosis was stage IVA gastric cancer. Pancreatoduodenectomy with portal vein resection could be important at this point. However, preoperative chemotherapy with S-1 and oxaliplatin was administered instead of performing extended surgery because the patient had poor general condition (performance status score of 3). The patient received three cycles of preoperative chemotherapy at the hospital along with rehabilitation and nutritional management with oral nutritional supplements. After treatment, the performance status score of the patient improved from 3 to 1. Furthermore, in terms of clinical therapeutic effect, the patient achieved partial response. Hence, laparoscopic distal gastrectomy with D2 lymph node dissection and partial transverse colectomy was performed. After surgery, the patient was admitted for oral intake on postoperative day 6 and was discharged on postoperative day 21. Based on the histopathological examination, gastric cancer had disappeared, and there were no evident malignant findings. Therefore, gastric cancer was classified as grade 3 according to the histological treatment efficacy criteria. The patient did not present with recurrence at 2 years after surgery. CONCLUSIONS: By actively administering preoperative chemotherapy, minimally invasive radical surgery with maximum preservation of the surrounding organs can be performed for locally far advanced-stage gastric cancer in older patients with poor general condition.

Idiopathic dendriform pulmonary ossification as the phenotype of interstitial lung abnormalities: CT-pathologic correlation and prevalence.

BACKGROUND AND PURPOSE: Idiopathic dendriform pulmonary ossification (DPO) is mostly asymptomatic, and detected incidentally in lung CT. There have been no reports on the precise CT-pathologic correlation and the prevalence of idiopathic DPO. This study aimed to clarify the histological background and prevalence of idiopathic DPO. MATERIALS AND METHODS: Sixteen patients with histologically confirmed idiopathic DPO (12 men and 4 women; mean age, 38.8 years; range 22-56 years) were identified in a nationwide epidemiological survey. Local HRCT findings of pre-biopsy examinations, such as branching, round, linear structures with or without high attenuation were compared side by side with histological findings. The attenuation of branching, round, and linear structures was classified into three-point levels on bone window images (width, 2500 HU; level, 500 HU). Furthermore, we collected continuous pulmonary CT images of 8111 cases for checking up metastasis from extrathoracic malignancy at a single institution, and evaluated the prevalence of interstitial lung abnormalities (ILAs) and DPO. RESULTS: In all 16 cases, branching (n = 15, 93%), round (n = 5, 31%), or linear (n = 5, 31%) structures were identified, histologically corresponding to dendriform ossification and cicatricial organizing pneumonia (OP)/fibrosis. Histologically, ossification was confirmed in all the 16 patients. However, in two cases, a highly attenuated structure could not be detected on the pre-biopsy CT of the same area. Regarding the prevalence of idiopathic DPO, 283 (3.5%) of 8111 patients had ILAs, of which a total of 26 (0.3% of all cases, 9.2% of ILAs cases) had DPO. CONCLUSION: Idiopathic DPO showed linear or branching structures with or without high attenuation on CT, corresponded to ossification, cicatricial OP/fibrosis. DPO was seen in 9.2% of ILAs cases. Idiopathic DPO is one of pathologic phenotypes of ILAs.

Lung miliary micro-nodules in human T-cell leukemia virus type I carriers.

Human T-cell leukemia virus type 1 (HTLV-1) carriers are rarely subject to inflammatory disorders in multiple organs, other than the well-known complication, adult T-cell leukemia/lymphoma (ATLL). HTLV-1 associated bronchiolo-alveolar disorder (HABA) has been proposed as an immune mediated pulmonary reaction seen rarely in HTLV-1 carriers. The reported clinico-pathological patterns of HABA are diffuse panbronchiolitis (DPB) and lymphoid interstitial pneumonia (LIP). We here report three cases of HTLV-1 carriers showing miliary micro-nodules throughout both lungs. Microscopic examination in the video assisted thoracic surgery biopsies demonstrated that all cases had multiple discrete micro-nodules which consisted of marked lymphoid infiltration, granulomas, eosinophils and a few foci of necrosis inside the granuloma. No findings indicating ATLL, other neoplastic conditions, infection or interstitial pneumonia, including DPB and LIP, were present following panels of special staining and immunohistochemical examinations. Two patients improved without treatment within one month, with no evidence of recurrence after 7 years. One patient showed slow deterioration of lung reticular shadows in spite of a low dose corticosteroid therapy (prednisolone 10 mg/day). We believe these cases may be a newly recognized variant of HABA.

Differential distribution of lymphatic clearance between upper and lower regions of the lung.

BACKGROUND AND OBJECTIVE: Many lung diseases arise as the consequence of inhalational injury. When pathogenic materials are inhaled, it is possible that their clearance routes become the main focus of injury in the lung. Lymphatic clearance is important in the removal from the lung of small inhaled particles. The leak of toxic agents from the lymphatic flow potentially explains the topographic distribution of diffuse lung diseases triggered by inhaled materials, for example asbestosis. The aim of this study was to evaluate the differences in lymphatic distribution across various craniocaudal levels of the lung by using carbon dust deposition (CDD) as a tracing marker. METHODS: We evaluated 61 HE-stained slides of the normal lung area from 61 lobectomy specimens. The intensity of CDD was scored into five degrees (0-4) in bronchiolovascular (BV) areas and subpleural/septal (SP) areas per slide. The distribution was highlighted by subtraction scores of SP areas from those of bronchiolovascular areas. RESULTS: The subtraction scores of the upper lung area were significantly greater than those in the lower area, whereas there was no significant difference in scores by pulmonary lobes. CONCLUSIONS: Our results indicated that the main lymphatic clearance routes vary according to the craniocaudal levels, and are predominant in BV areas of the upper areas and in SP areas of the lower areas of the lung. This may explain the histological variations in anatomical distribution observed in the biopsy specimens of some diffuse lung diseases.

Diagnostic yield and the number of tumor cells of ultrathin bronchoscopy for peripheral lung lesions: A comparison with thin bronchoscopy.

Ultrathin bronchoscopy has been reported to have a higher diagnostic yield than thin bronchoscopy for small peripheral lung lesions in transbronchial biopsy under radial endobronchial ultrasonography (EBUS). However, data comparing the number of tumor cells in non-small cell lung cancer (NSCLC) are limited. We retrospectively compared the number of NSCLC tumor cells in peripheral lung lesions obtained using an ultrathin bronchoscope and a thin bronchoscope with radial EBUS between April 2020 and October 2021. In all patients, we used virtual bronchoscopic navigation (VBN) software, and guide sheaths were used in thin bronchoscopy cases. A total of 175 patients were enrolled in this study. Ultrathin bronchoscopy cases (n = 69) had lesions with a smaller diameter that are more peripherally located compared to thin bronchoscopy cases (n = 106) (median, 25.0 vs. 26.5 mm, mean bronchial generations accessed by bronchoscopy; 4.4±1.2 vs. 3.8±1.0, respectively; p<0.010). There were no significant differences in the overall diagnostic yield (ultrathin vs. thin bronchoscopy cases, 68.1% vs. 72.6%, p = 0.610) or diagnostic yield in only lung cancer cases (78.6% vs. 78.5%, p = 1.000). In histologically NSCLC cases (n = 102), the maximum number of tumor cells per slide as the primary endpoint was similar (average, 307.6±246.7 vs. 328.7±314.9, p = 0.710). The success rate of the Oncomine™ analysis did not differ significantly (80.0% vs. 55.6%, p = 0.247). The yield of NSCLC tumor cells was not different between the samples obtained by the ultrathin bronchoscope and those obtained by the thin bronchoscope.

Overcoming the Interobserver Variability in Lung Adenocarcinoma Subtyping: A Clustering Approach to Establish a Ground Truth for Downstream Applications.

CONTEXT.­: The accurate identification of different lung adenocarcinoma histologic subtypes is important for determining prognosis but can be challenging because of overlaps in the diagnostic features, leading to considerable interobserver variability. OBJECTIVE.­: To provide an overview of the diagnostic agreement for lung adenocarcinoma subtypes among pathologists and to create a ground truth using the clustering approach for downstream computational applications. DESIGN.­: Three sets of lung adenocarcinoma histologic images with different evaluation levels (small patches, areas with relatively uniform histology, and whole slide images) were reviewed by 17 international expert lung pathologists and 1 pathologist in training. Each image was classified into one or several lung adenocarcinoma subtypes. RESULTS.­: Among the 4702 patches of the first set, 1742 (37%) had an overall consensus among all pathologists. The overall Fleiss κ score for the agreement of all subtypes was 0.58. Using cluster analysis, pathologists were hierarchically grouped into 2 clusters, with κ scores of 0.588 and 0.563 in clusters 1 and 2, respectively. Similar results were obtained for the second and third sets, with fair-to-moderate agreements. Patches from the first 2 sets that obtained the consensus of the 18 pathologists were retrieved to form consensus patches and were regarded as the ground truth of lung adenocarcinoma subtypes. CONCLUSIONS.­: Our observations highlight discrepancies among experts when assessing lung adenocarcinoma subtypes. However, a subsequent number of consensus patches could be retrieved from each cluster, which can be used as ground truth for the downstream computational pathology applications, with minimal influence from interobserver variability.

Yield of tumor samples with a large guide-sheath in endobronchial ultrasound transbronchial biopsy for non-small cell lung cancer: A prospective study.

BACKGROUND: Adequate tumor tissue is required to make the best treatment choice for non-small cell lung cancer (NSCLC). Transbronchial biopsy (TBB) by endobronchial ultrasonography with a guide sheath (EBUS-GS) is useful to diagnose peripheral lung lesions. The data of tumor cell numbers obtained by two different sizes of GSs is limited. We conducted this study to investigate the utility of a large GS kit to obtain many tumor cells in patients with NSCLC. METHODS: Patients with a peripheral lung lesion and suspected of NSCLC were prospectively enrolled. They underwent TBB with a 5.9-mm diameter bronchoscope with a large GS. When the lesion was invisible in EBUS, we changed to a thinner bronchoscope and TBB was performed with a small GS. We compared the tumor cell number prospectively obtained with a large GS (prospective large GS group) and those previously obtained with a small GS (small GS cohort). The primary endpoint was the tumor cell number per sample, and we assessed characteristics of lesions that could be obtained by TBB with large GS. RESULTS: Biopsy with large GS was performed in 55 of 87 patients (63.2%), and 37 were diagnosed with NSCLC based on histological samples. The number of tumor cells per sample was not different between two groups (658±553 vs. 532±526, estimated difference between two groups with 95% confidence interval (CI); 125 (-125-376), p = 0.32). The sample size of the large GS group was significantly larger than that of the small GS cohort (1.75 mm2 vs. 0.83 mm2, estimated difference with 95% CI; 0.92 (0.60-1.23) mm2, p = 0.00000019). The lesion involving a third or less bronchus generation was predictive factors using large GS. CONCLUSIONS: The sample size obtained with large GS was significantly larger compared to that obtained with small GS, but there was no significant difference in tumor cell number. The 5.9-mm diameter bronchoscope with large GS can be used for lesions involving a third or less bronchus generation.

Association between neurosarcoidosis with autonomic dysfunction and anti-ganglionic acetylcholine receptor antibodies.

OBJECTIVE: To determine whether autonomic dysfunction in neurosarcoidosis is associated with anti-ganglionic acetylcholine receptor (gAChR) antibodies, which are detected in autoimmune autonomic ganglionopathy. METHODS: We retrospectively extracted cases of sarcoidosis from 1787 serum samples of 1,381 patients between 2012 and 2018. Anti-gAChR antibodies against the α3 and ß4 subunit were measured by luciferase immunoprecipitation to confirm the clinical features of each case. We summarized literature reviews of neurosarcoidosis with severe dysautonomia to identify relevant clinical features and outcomes. RESULTS: We extracted three new cases of neurosarcoidosis with severe dysautonomia, among which two were positive for anti-gAChR antibodies: Case 1 was positive for antibodies against the ß4 subunit, and Case 2 was positive for antibodies against both the α3 and ß4 subunits. We reviewed the cases of 15 patients with neurosarcoidosis and severe dysautonomia, including the three cases presented herein. Orthostatic hypotension and orthostatic intolerance were the most common symptoms. Among the various types of neuropathy, small fiber neuropathy (SFN) was the most prevalent, with seven of nine cases exhibiting definite SFN. Six of eight cases had impaired postganglionic fibers, of which the present three cases revealed abnormality of 123I-MIBG myocardial scintigraphy. Of the 11 cases, 10 were responsive to immunotherapy, except one seropositive case (Case 2). CONCLUSIONS: The presence of gAChR antibodies may constitute one of the mechanisms by which dysautonomia arises in neurosarcoidosis.

Clinical, radiological, and pathological evaluation of "NSIP with OP overlap" pattern compared with NSIP in patients with idiopathic interstitial pneumonias.

BACKGROUND: Nonspecific interstitial pneumonia (NSIP) and organizing pneumonia (OP) are major subtypes of idiopathic interstitial pneumonias (IIPs) and closely related to connective tissue diseases (CTDs). "NSIP with OP overlap" is a controversial finding that has recently appeared in the criteria of interstitial pneumonia with autoimmune features (IPAF). However, details of this controversial entity are not well known. OBJECTIVE: To determine the frequency of "NSIP with OP overlap" pattern in IIPs and to identify differences from idiopathic NSIP (iNSIP). METHODS: In 524 patients with interstitial pneumonia from 39 institutes who underwent surgical lung biopsy, 444 were diagnosed as IIPs by a multidisciplinary discussion meeting via a cloud-based integrated database. Among these patients, 44 (9.9%) who had iNSIP and 21 (4.7%) with histopathologically-defined "NSIP with OP overlap" pattern (a pathological NSIP and OP pattern, but without a UIP pattern) were retrospectively studied. RESULTS: Patients with "NSIP with OP overlap" pattern showed a significantly greater extent of consolidation (p < 0.001), more subpleural ground glass attenuation (p = 0.036), and more peripheral + bronchovascular distribution (p = 0.009) on high-resolution computed tomography than those with iNSIP. The incidences of newly-developed CTDs during follow-up was similar between the groups and polymyositis/dermatomyositis was the most frequent CTD in both groups. Nearly half of the patients fulfilled IPAF criteria, but no significant difference was found between iNSIP and "NSIP with OP overlap" pattern (47.7% vs. 42.9, p = 0.712). The incidence of acute exacerbation and the survival rates were similar between the groups. CONCLUSIONS: The incidence of "NSIP with OP overlap" pattern is 4.7% in IIPs. The frequency of newly-developed CTDs during follow-up, mainly polymyositis/dermatomyositis, the frequency of acute exacerbation, and the survival rate in "NSIP with OP overlap" pattern are similar to those of iNSIP.