According to the lung cancer staging project, T1a (≤ 2 cm) non-small-cell lung cancer (NSCLC) should be additionally classified into ≤ 1 cm and > 1 to 2 cm groups. This study aimed to investigate the ...surgical procedure for NSCLC ≤ 1 cm and > 1 to 2 cm.
We identified 15,760 patients with T1aN0M0 NSCLC after surgery from the Surveillance, Epidemiology, and End Results database. Overall survival (OS) and lung cancer-specific survival (LCSS) were compared among patients after lobectomy, segmentectomy, or wedge resection. The proportional hazards model was applied to evaluate multiple prognostic factors.
OS and LCSS favored lobectomy compared with segmentectomy or wedge resection in patients with NSCLC ≤ 1 cm and > 1 to 2 cm. Multivariable analysis showed that segmentectomy and wedge resection were independently associated with poorer OS and LCSS than lobectomy for NSCLC ≤ 1 cm and > 1 to 2 cm. With sublobar resection, lower OS and LCSS emerged for NSCLC > 1 to 2 cm after wedge resection, whereas similar survivals were observed for NSCLC ≤ 1 cm. Multivariable analyses showed that wedge resection is an independent risk factor of survival for NSCLC > 1 to 2 cm but not for NSCLC ≤ 1 cm.
Lobectomy showed better survival than sublobar resection for patients with NSCLC ≤ 1 cm and > 1 to 2 cm. For patients in whom lobectomy is unsuitable, segmentectomy should be recommended for NSCLC > 1 to 2 cm, whereas surgeons could rely on surgical skills and the patient profile to decide between segmentectomy and wedge resection for NSCLC ≤ 1 cm.
Importance There is a lack of studies exploring the performance of a deep learning survival neural network in non–small cell lung cancer (NSCLC). Objectives To compare the performances of DeepSurv, a ...deep learning survival neural network with a tumor, node, and metastasis staging system in the prediction of survival and test the reliability of individual treatment recommendations provided by the deep learning survival neural network. Design, Setting, and Participants In this population-based cohort study, a deep learning–based algorithm was developed and validated using consecutive cases of newly diagnosed stages I to IV NSCLC between January 2010 and December 2015 in a Surveillance, Epidemiology, and End Results database. A total of 127 features, including patient characteristics, tumor stage, and treatment strategies, were assessed for analysis. The algorithm was externally validated on an independent test cohort, comprising 1182 patients with stage I to III NSCLC diagnosed between January 2009 and December 2013 in Shanghai Pulmonary Hospital. Analysis began January 2018 and ended June 2019. Main Outcomes and Measures The deep learning survival neural network model was compared with the tumor, node, and metastasis staging system for lung cancer–specific survival. The C statistic was used to assess the performance of models. A user-friendly interface was provided to facilitate the survival predictions and treatment recommendations of the deep learning survival neural network model. Results Of 17 322 patients with NSCLC included in the study, 13 361 (77.1%) were white and the median (interquartile range) age was 68 (61-74) years. The majority of tumors were stage I disease (10 273 59.3%) and adenocarcinoma (11 985 69.2%). The median (interquartile range) follow-up time was 24 (10-43) months. There were 3119 patients who had lung cancer–related death during the follow-up period. The deep learning survival neural network model showed more promising results in the prediction of lung cancer–specific survival than the tumor, node, and metastasis stage on the test data set (C statistic = 0.739 vs 0.706). The population who received the recommended treatments had superior survival rates than those who received treatments not recommended (hazard ratio, 2.99; 95% CI, 2.49-3.59;P < .001), which was verified by propensity score–matched groups. The deep learning survival neural network model visualization was realized by a user-friendly graphic interface. Conclusions and Relevance The deep learning survival neural network model shows potential benefits in prognostic evaluation and treatment recommendation with respect to lung cancer–specific survival. This novel analytical approach may provide reliable individual survival information and treatment recommendations.
When multiple target lung nodules exist, the computed tomography (CT)-guided percutaneous localization procedure becomes complicated. In this study, a three-dimensional (3D)-printed template was ...designed that could guide hook wire localization of multiple lung nodules. The pilot study aimed for preliminary validation of the feasibility of template-guided localization for multiple lesions.
Patients with multiple lung nodules (<2 cm) and who were scheduled for surgical resection were recruited for participation in this study. After securing their preadmission CT images, the study investigators reconstructed a 3D thorax model from which they designed a digital model as a navigational template. A physical template was then printed for guiding the percutaneous localization of lung nodules. The localization accuracy was evaluated on the basis of the deviation between the localizer and the nodule.
From April 2018 to November 2018, the study enrolled 16 patients with 34 lung nodules. All nodules were successfully localized under template guidance, with a median procedural time of 10.0minutes (interquartile range IQR, 8.5-12.6 minutes) and a median radiation exposure of 235 mGy • cm (IQR, 195-254 mGy • cm). The median deviation from the hook wires and nodule centers was 9.0 mm (IQR, 6.2-11.8 mm). Except for 2 cases of pneumothorax without need for further intervention, no complications occurred.
Navigational templates built using 3D printing may serve as an effective approach for facilitating localization of multiple lung nodules.
Patients in the intensive care unit (ICU) are often in critical condition and have a high mortality rate. Accurately predicting the survival probability of ICU patients is beneficial to timely care ...and prioritizing medical resources to improve the overall patient population survival. Models developed by deep learning (DL) algorithms show good performance on many models. However, few DL algorithms have been validated in the dimension of survival time or compared with traditional algorithms.
Variables from the Early Warning Score, Sequential Organ Failure Assessment Score, Simplified Acute Physiology Score II, Acute Physiology and Chronic Health Evaluation (APACHE) II, and APACHE IV models were selected for model development. The Cox regression, random survival forest (RSF), and DL methods were used to develop prediction models for the survival probability of ICU patients. The prediction performance was independently evaluated in the MIMIC-III Clinical Database (MIMIC-III), the eICU Collaborative Research Database (eICU), and Shanghai Pulmonary Hospital Database (SPH).
Forty variables were collected in total for model development. 83 943 participants from 3 databases were included in the study. The New-DL model accurately stratified patients into different survival probability groups with a C-index of >0.7 in the MIMIC-III, eICU, and SPH, performing better than the other models. The calibration curves of the models at 3 and 10 days indicated that the prediction performance was good. A user-friendly interface was developed to enable the model's convenience.
Compared with traditional algorithms, DL algorithms are more accurate in predicting the survival probability during ICU hospitalization. This novel model can provide reliable, individualized survival probability prediction.
Micropapillary (MIP) component was a major concern in determining surgical strategy in lung adenocarcinoma (LUAD). We sought to develop a novel method for detecting MIP component during surgery.
...Differentially expressed proteins between MIP-positive and MIP-negative LUAD were identified through proteomics analysis. The semi-dry dot-blot (SDB) method which visualises the targeted protein was developed to detect MIP component.
Cellular retinoic acid-binding protein 2 (CRABP2) was significantly upregulated in MIP-positive LUAD (P < 0.001), and the high CRABP2 expression zone showed spatial consistency with MIP component. CRABP2 expression was also associated with decreased recurrence-free survival (P < 0.001). In the prospective cohort, the accuracy and sensitivity of detecting MIP component using SDB method by visualising CRABP2 were 82.2% and 72.7%, which were comparable to these of pathologist. Pathologist with the aid of SDB method would improve greatly in diagnostic accuracy (86.4%) and sensitivity (78.2%). In patients with minor MIP component (≤5%), the sensitivity of SDB method (63.6%) was significantly higher than pathologist (45.4%).
Intraoperative examination of CRABP2 using SDB method to detect MIP component reached comparable performance to pathologist, and SDB method had notable superiority than pathologist in detecting minor MIP component.
This study evaluated the clinicopathologic characteristics and prognostic impact of atypical epidermal growth factor receptor (EGFR) mutations in patients with completely resected lung adenocarcinoma ...(LUAD) and investigate whether adjuvant chemotherapy could benefit the survival outcomes for these subjects.
We retrospectively reviewed resected LUAD samples from 8437 patients and identified 5358 EGFR-mutated (EGFRm) cases. Of these, 4847 had classical mutations, while 511 had atypical mutations. For further survival analysis, propensity score matching, Kaplan–Meier curve, and Cox regression analyses were conducted.
Of the 511 patients with atypical EGFRm LUAD, 131 patients had compound mutations. The frequency of exon 20 insertion (20-ins), G719X, L861Q, S768I, and de novo T790M were 30.3%, 32.7%, 21.9%, 9.2%, and 11.4%, respectively. These patients included a higher proportion of males than those with classical EGFRm LUAD. Between the 483 matched pairs of the classical and atypical EGFRm patients, no significant difference emerged in disease-free survival (DFS) (p = 0.476). Patients with the L861Q mutation had the poorest DFS among those with atypical EGFRm LUAD (p = 0.005). Cox regression analyses revealed that the L861Q mutation was an independent prognostic factor for DFS in 487 patients with solely atypical EGFRm LUAD. In addition, adjuvant chemotherapy did not improve the DFS for those patients, whether in stage IB (p = 0.638) or II-III (p = 0.505) of the disease.
The L861Q mutation is an independent prognostic factor for DFS in patients with atypical EGFRm LUAD after complete resection who would not benefit from adjuvant chemotherapy regardless of disease stage.
Display omitted
•Resected atypical and classical EGFR-mutated lung adenocarcinoma have similar DFS.•L861Q mutation is associated with the poorest DFS among atypical EGFR mutations.•Adjuvant chemotherapy couldn't benefit atypical EGFR-mutated lung adenocarcinoma.
Patients with NSCLC with M1a disease regardless of lymph node status were categorized as stage IV. This study aims to investigate whether the N descriptors in M1a patients could provide clinical ...information.
Overall, 39,731 patients with NSCLC with M1a disease were identified from the Surveillance, Epidemiology, and End Results database during 2005–2012. Lung cancer–specific survival (LCSS) was compared among M1a patients stratified by N stage. A Cox proportional hazards regression model was applied to evaluate the prognostic factors. Statistical analyses were performed in all subgroups.
M1a patients without lymph node involvement had the best LCSS, followed by patients with N1 disease; no difference in LCSS was observed between N2 and N3 disease (N0 versus N1, p < 0.001; N1 versus N2, p < 0.001; and N2 versus N3, p = 0.478). Similarly, this trend was observed when patients were subdivided into two temporal cohorts (2005–2008 and 2009–2012) and also when M1a disease was subdivided into contralateral pulmonary nodules and pleural dissemination (malignant pleural effusion or pericardial effusion and pleural nodules). In addition, a difference in LCSS between N2 and N3 disease was observed in patients with malignant pleural nodules (p = 0.003). Multivariate analysis showed that lymph node involvement was an independent prognostic factor for M1a patients, and this result was also noticed in all subgroups.
These results provide preliminary evidence that lymph node stage may have clinical significance among patients with NSCLC with M1a disease, adding prognostic information.
Background
This study aimed to validate the R classification including uncertain resection (R‐un) proposed by the International Association for the Study of Lung Cancer (IASLC) in a Chinese non‐small ...cell lung cancer (NSCLC) population.
Methods
The study retrospectively investigated a 2009–2013 single‐institutional NSCLC resection cohort in China. After reclassification, recurrence‐free survival (RFS) and overall survival (OS) were calculated using survival analyses and compared with those using the 2005 version IASLC R classification.
Results
Under the proposed stratification, 3819 (72.1%) individuals were classified as R0, 1371 (25.9%) as R‐un, 71 (1.3%) as R1, and 32 (0.6%) as R2. The 5‐year OS probabilities for the R0, R‐un, and R1/R2 groups were 71%, 46%, and 34%, respectively. The prognostic stratification remained significant in the fully adjusted Cox models (p < 0.001). Compared with the original classification, Harrell's concordance index of reclassification improved significantly, from 0.508 to 0.679 for RFS and from 0.510 to 0.692 for OS (RFS: p = 0.007; OS: p = 0.001). The survival analysis showed that R‐un patients with highest mediastinal lymph node station metastasis had significantly worse survival than R0 patients with mediastinal nodal metastasis (RFS: 44 vs. 36 months, hazard ratio HR: 1.29, p < 0.001; OS: 59 vs. 50 months, HR: 1.34, p < 0.001). Cox proportional hazards regression analysis showed that highest mediastinal lymph node station metastasis was an independent risk factor for RFS (HR: 1.22) and OS (HR: 1.25).
Conclusions
The proposed R classification showed valid prognostic stratification, including highest mediastinal nodal station metastasis.
The International Association for the Study of Lung Cancer (IASLC) proposed a revised R classification to upstage extracapsular extension (ECE) of tumor in nodes from R0 to R1. Nevertheless, evidence ...to confirm this proposal is insufficient.
The study included 4061 surgical patients with NSCLC. After reclassification by IASLC-R classification, overall survival (OS) was analyzed to compare patients with ECE with those with R0, R(un), and incomplete resection (R1 and R2). The recurrence pattern of ECE was evaluated to determine whether it correlated with incomplete resection.
Among 1136 patients with N disease, those without ECE (n = 754, 67%) had a significantly better OS than those with ECE (n = 382, 33%) (p < 0.001). This negative prognostic significance was consistent across multiple subgroups. Multivariate analysis revealed that ECE was an independent prognostic risk factor (p < 0.001). When patients with ECE were separated from the IASLC-R1 group, their OS was significantly worse than that of IASLC-R(un) patients, but comparable to that of the remaining patients in the IASLC-R1 patients when analyzing all patients and patients with N disease. Moreover, patients with ECE had an increased risk of local recurrence in the mediastinum (p < 0.001), ipsilateral lung (p = 0.031), and malignant pleural effusion or nodes (p = 0.004) but not distant recurrence including contralateral or both lungs (p = 0.268), liver (p = 0.728), brain (p = 0.252), or bone (p = 0.322).
The prognosis of ECE patients is comparable with that of R1 patients. Moreover, their higher risk of local recurrence strongly suggests the presence of occult residual tumor cells in the surgical hemithoracic cavity. Therefore, upgrading ECE into incomplete resection is reasonable.